1. This site uses cookies. By continuing to use this site, you are agreeing to our use of cookies. Learn More.
  2. Dear Guest, Please note that adult content is not permitted on this forum. We have had our Google ads disabled due to some posts that were found from some time ago. Please do not post adult content and if you see any already on the forum, please report the post so that we can deal with it. Thanks, scara
    Dismiss Notice
  3. Please read this post: https://www.glory-glory.co.uk/community/threads/fund-raising-for-jeremy-yermiyahu.10542/
    Dismiss Notice

Nuno Espírito Santo

Discussion in 'Spurs News & Views' started by Denmark Spur, 30 Jun 2021.

  1. Bedfordspurs

    Bedfordspurs Mel Hopkins

    Joined:
    14 Jan 2012
    Messages:
    32,960
    Likes Received:
    21,661
    Saves
    I talked about keepers making saves
    Watford’s keeper made the most saves of any keeper in the league this season. That shows a dominance of something surely?
    How is that not key ???
    Hugo didn’t in any of the games
    Vs Wolves their keeper made as many saves as Hugo but they were better saves than Hugo’s mainly as our shots on target had a Higher xG (if we’re using that stat)
    quality of shots or shots making the keeper work is key to winning games
    I’ve said it before as an issue with xG… a team taking 20 shots from distance at 0.1Xg doesn’t mean they will score 2 goals does it. Cumulative xG is a bit do farce in reality
    What’s key is the actual highest relevant part of that xG score. That’s the golden chance to score and your best effort

    I am more than happy for teams to keep on taking low quality pot shots. As long as we take out chances we will get the points

    palace we don’t create any chances and we conceded goals when we were down to 10
     
  2. braineclipse

    braineclipse Jimmy Cantrell

    Joined:
    15 Jan 2012
    Messages:
    19,493
    Likes Received:
    7,457
    Quite certain I saw a stat before the Palace game that Lloris had made the most saves of any PL goalkeeper at that point this season. It's not the same now, but was at the time. If the Watford keeper making saves tells us something it should tell us the same about us conceding shots, surely. Particularly as long as both are from low xG chances.

    20 shots with an xG of 0.1 doesn't mean the team will score two goals. Over time though, if you keep conceding those shots, you will concede whatever the keeper can't make up for. That's how xG is created, looking at the likelihood of shots going in from various areas. All else being equal, on average you'll concede twice with an xG of 2 against you.

    Everything else isn't equal though, as you point out. Lloris will outperform xG against over time, he's clearly shown that. To what extent though?

    We may be better at getting blocks in, getting closer to attackers to put them off balance, rush their shots etc. I'll believe it at a later stage if it keeps happening. For now random chance seems just as likely for that.
     
    Finney Is Back likes this.
  3. Bedfordspurs

    Bedfordspurs Mel Hopkins

    Joined:
    14 Jan 2012
    Messages:
    32,960
    Likes Received:
    21,661
    I was referring to the stat as saves in one game
    No idea cumulatively. It’s part of the issue when you don’t take each game in isolation

    when you look at the xG data the average quality of chance we concede is very very low

    and I agree that teams spanking lots of crap shots doesn’t equal us conceding. But if there that bad at shooting it’s reasonable to expect the keeper to save them. It also comes down to who is taking those shots which xG doesn’t factor in. Watford’s best chance fell to sissoko according to xG… I’d fancy Hugo to do better in front of goal than him (shot was blocked)

    Why would an improvement in defending basics be seen as random? I mean with 11 men vs palace we were still defending really well, just creating nothing. Their goal that changed the game came from an individual error from a poor player making a handball.

    We crumbled defensively for two reasons last season; fitness and individual errors (probably both tied together). We saw on Saturday how an individual error cost us. Nuno I believe has been working hard on both those aspects but it takes time
     
    StephenH likes this.
  4. Finney Is Back

    Finney Is Back Paul Walsh

    Joined:
    26 Jan 2012
    Messages:
    12,052
    Likes Received:
    9,856
    So for you it makes a difference if the shot happens to be on target or not but you don't feel that there is any relevance how good the chance actually is (i.e. the XG). You are picking and choosing stats to suit your argument. A shot from the 8 yards out in the middle of the goal with the player clean through and under no pressure that happens to be put wide so the keeper doesn't have to make a save is statistically a FAR better chance than a shot from 25 yards out with several defenders between the player and the goal and the player under pressure... If the player then happens to get his shot on target so that the keeper makes a save that doesn't mean it was a better chance than the one the keeper didn't have to save.

    Actually a team taking 20 shots with an average XG of 0.1 absolutely means that they are likely to score two goals. That's exactly how XG is worked out. Of course in one particular isolated match it doesn't mean that will be the case but in all matches over the all seasons the average goals scored from 20 x 0.1XG chances is 2. That's how the chance is given an XG of 0.1... because history shows that 1 in 10 of those chances are scored.

    Interestingly enough in the BBC write up of the Watford game the Watford keeper was rated Watford's 11th best player with a rating of 5.34. That certainly doesn't suggest to me that their keeper kept them in the game!
     
    Last edited: 15 Sep 2021
  5. Bedfordspurs

    Bedfordspurs Mel Hopkins

    Joined:
    14 Jan 2012
    Messages:
    32,960
    Likes Received:
    21,661
    XG assesses each shot
    You can’t use it cumulative
    A team has one shot with an xG of 1 and scores
    Has done better than a team with 15 shots with an xG of 0.1 each and not scored
    It looks at each shot in isolation

    If a team takes plenty of shots at 0.1…. That means their shots are likely to go in 10% of the time. You can’t add 10 shots to get a goal as their all still the same crap shot. Unless football fans think 10% of nothing multiplied by 10 should in theory mean something

    and bbc ratings are fans rating don’t forget
    The guy met in a free kick that wasn’t a shot
     
    Last edited: 15 Sep 2021
  6. DubSpur

    DubSpur Neil Ruddock

    Joined:
    28 Aug 2015
    Messages:
    1,929
    Likes Received:
    2,477
    You can use it cumulatively. It's expected goals not actual goals in a single game. It's like anything in statistics, it'll lean towards that over the long term but there will be short term anomalies. If your Xg was 200 over two seasons it would end up being fairly accurate, just like if you toss a coin 10k times it'll be close to 5k heads and 5k tails.
    In one game you might have an Xg of 5 and score 1, just like if you toss 10 coins you might get 8 heads and two tails. It's a guide to how you will do long term.
     
    Bullet and Finney Is Back like this.
  7. Bedfordspurs

    Bedfordspurs Mel Hopkins

    Joined:
    14 Jan 2012
    Messages:
    32,960
    Likes Received:
    21,661
    But it’s measured short term. It’s measured per shot. The actual data is singular so adding it in a game like football doenst work
    So cumulative the whole thing fails
    It needs a variable due to quantity
    You use the coin toss example… it’s 50-50 but as you say you could flip and get one side 80% of the time. Your odds are still 50-50 for each toss
    So the frequency doenst change the likelihood. And in sport, the more you repeat something the easier it becomes to stop
     
  8. Finney Is Back

    Finney Is Back Paul Walsh

    Joined:
    26 Jan 2012
    Messages:
    12,052
    Likes Received:
    9,856
    I'm not sure there is such a thing as a shot with an XG of 1 to be honest.... perhaps middle of the goal, one yard out when the player is beyond the keeper?

    But to use a more likely example.... Yes a team who has 15 x 0.1XG changes is likely to do better than a team who has 2 x 0.5XG chances. It is the historical statistics that absolutely prove that (i.e. where the XG comes from in the first place). Do you understand how statistics work?

    To your point about the BBC ratings.... Yes the fans give those ratings. Therefore the Watford and Spurs fans (the fans most likely to have voted on that game) feel that the Watford keeper was the 11th best player for Watford on that day. I don't think that bears out your view that he had a good game because he made 8 saves.
     
  9. DubSpur

    DubSpur Neil Ruddock

    Joined:
    28 Aug 2015
    Messages:
    1,929
    Likes Received:
    2,477
    Yes, and if you shoot with an Xg goal of 0.5 your odds are 50-50 for each shot...
     
  10. Bedfordspurs

    Bedfordspurs Mel Hopkins

    Joined:
    14 Jan 2012
    Messages:
    32,960
    Likes Received:
    21,661
    Agree
    Each shot
    So the shot in isolation
    Hence the issue cumulative adding up the shots
     
  11. Bedfordspurs

    Bedfordspurs Mel Hopkins

    Joined:
    14 Jan 2012
    Messages:
    32,960
    Likes Received:
    21,661
    I under stand statistics
    When I was at Uni I studied them as part fo engineering
    But… you also have to include variables and conditions for them to have any meaning
    And in sport, particularly football you have a LOT of variables
    its why for simple chance analysis its fine. For cumulative data it isnt IMO
     
  12. Finney Is Back

    Finney Is Back Paul Walsh

    Joined:
    26 Jan 2012
    Messages:
    12,052
    Likes Received:
    9,856
    If I have a standard dice I have a 1/6 chance of getting a six (or any other specific number).

    If I have a standard coin I have a 1/2 chance of getting a head (or a tail)

    If I roll the dice 20 times and toss the coin 3 times am I statistically more likely to more heads or more sixes?
     
  13. Bedfordspurs

    Bedfordspurs Mel Hopkins

    Joined:
    14 Jan 2012
    Messages:
    32,960
    Likes Received:
    21,661
    What are your variables?
    Where’s it landing
    What time of day/the match
    What weather
    Who is stopping it landing or rolling
    Hence my comment
    It’s useless without those variables unless it’s analyse for each individual chance
    Tossing a coin 50/50 - we agree
    Rolling a die - 1 in 6 chance, I agree
    Each chance is unique
    Cumulatively I’m you have data, but the likelihood doesn’t change
     
  14. Finney Is Back

    Finney Is Back Paul Walsh

    Joined:
    26 Jan 2012
    Messages:
    12,052
    Likes Received:
    9,856
    I think you need to take a look at how XG is calculated and the variables that go into the calculations. It is not an exact science but getting better and better as more and more data goes in. Take a look at the XG table for last season and it was pretty damn good with only really Brighton being a statistical anomaly (XG position 5th, actual position 16th). I'm pretty sure that you were one of the advocates for us taking Potter as manager as well? (strange with you not believing in XG)?
     
  15. Bedfordspurs

    Bedfordspurs Mel Hopkins

    Joined:
    14 Jan 2012
    Messages:
    32,960
    Likes Received:
    21,661
    I like xG for each shot
    The fact there isn’t one standard model though is worrying
    I really really don’t like it for cumulative shots
    I know how it works
    I’ve read up on it as I’m interested in data
    As I’ve said it doesn’t have enough variables for anything other than analysing a shot in isolation because of that
    Would you agree a shot form Kane is likely to be better than a shot from sissoko? That’s not factored into any model I know of
    Timing of the effort is massive variable. Late in the game, tired legs mean better opportunities. Not factored in
    Home to away and crowds - I don’t think that’s in any model either
    The state of the pitch, ball, the opposition
    Hence why I don’t mind it for indivisible chances, but not cumulatively.
    And the Brighton xG was goals scored in a game (so expected points one). Didn’t factor crap strikers, better opposition etc…
    As I said vs Wolves we created the better chances. I saw it with my eyes, xG backed that up
    Same as Watford. I added into the Watford mix that their keeper made 7 saves which was the most of any keeper on the division at that point. The saves though were a combination of poor shot execution and good work from him. But he did massively rooster up sons shot
     
  16. Bedfordspurs

    Bedfordspurs Mel Hopkins

    Joined:
    14 Jan 2012
    Messages:
    32,960
    Likes Received:
    21,661
  17. Diamond Lights

    Diamond Lights Allan Nielsen

    Joined:
    18 Jan 2012
    Messages:
    1,737
    Likes Received:
    2,406
    Location:
    North Stand
    Agree but Levy needed to hire the best man for the job. If that was Potter, then so be it. Especially given the 'attacking' talk he mentioned. 3 DMs away to Palace says different!
     
  18. Bedfordspurs

    Bedfordspurs Mel Hopkins

    Joined:
    14 Jan 2012
    Messages:
    32,960
    Likes Received:
    21,661
    Potter really would have been brilliant for us
     
    metalgear and StephenH like this.
  19. AdamB

    AdamB Les Howe

    Joined:
    14 Jan 2012
    Messages:
    20,290
    Likes Received:
    9,721
    Correct. And 73% of statistics are made up on the spot
     
    metalgear and Bedfordspurs like this.
  20. Bedfordspurs

    Bedfordspurs Mel Hopkins

    Joined:
    14 Jan 2012
    Messages:
    32,960
    Likes Received:
    21,661
    What kind of spot?
     
    metalgear and DTA like this.

Share This Page