Val Ball and stats
#21
(02-03-2022, 09:09 AM)strawman Wrote:
(02-03-2022, 09:01 AM)TartanRug Wrote:
(02-03-2022, 08:26 AM)Derek Hardballs Wrote: “A manager who relies heavily on statistics, the Frenchman – like many in the modern game – is a fan of expected goals data, or xG as it is known.” Masi Express and Star

That explains the one dimensional, robotic, predictable tactics we employed. Relying on statistics like XFuggingG is not managing and it ignores what you can see with your own eyes. The stats said we were creating more chances than the opposition but that only tells a fraction of the story, most of those chances weren’t gilt edged they were half chances from poor crosses, corners, set pieces, balls lumped into the box and long throws. It’s no surprise it only took a handful of games for managers to work out how to beat us employing these ‘ball free’ tactics. 

Even on here some of us said as early as the Sheff Utd match if you just sit deep, play it long and don’t dawdle on the ball in your own half you have very workable tactics to negate any XG inspired tactics we employed. 

Sadly this is what I will remember from Val’s time with us. It’s a shame as I think he’s a decent person who got lost in the theory and it harmed his practice.

XG does account for the quality of chance you know. A half chance will hardly get any added to the Xg, chances like Dike's header v Peterboro would get quite a bit added. It tells you (roughly) how good your finishing is. 

Early in the season we were probably (I havent checked) matching or exceeding our Xg in games like Cardiff away, then came a period when we were missing a lot of good chances, then the chances started to dry up and in the last couple of games unsurprisingly we've been comfortably outscored in Xg

But not the quality of the footballer who has the chance - based on the 'average' and apparently all ours are below average  Cool

No exactly, our players were performing below the average in terms of finishing. I'm sure they practice finishing in training enough so not sure what else a manager can do when you're continually missing chances. When these started to dry up though Val needed to remedy it and he didn't.
Reply
#22
(02-03-2022, 09:12 AM)Ted Maul Wrote:
(02-03-2022, 09:02 AM)MassDebater Wrote: I agree with a lot of that, but surely Six would have really used the in depth player stats BEFORE signing Hugill? And he's never ever been an amazing goal scoring striker (over 4.5 games per goal, any stat based manager would think, hmmm maybe not, surely?)

Stats have a place without a doubt, but they should never be the be all and end all. There were several games when people on here were bigging up our xG, but the reality to me and others was that we still never looked like scoring. I think the people who run xG need to re-evaluate their algorithms to account for us this season, we've definitely been the epitome of a statistical anomaly.

There is not a chance in hell Hugill was his first pick of strikers available in the summer. If he was he should have been dismissed on the spot.

The issue with xG is that it's easily derided once you have seen the result. But as a longer term indicator or performance and likelihood to win matches it's a useful tool. A one off game where you miss hatfuls of chances and the opposition nick one you can live with, happens to everyone and it shouldn't make you rip up your notebook. Do you think Pep panics when he loses his annual home game against Palace/Wolves or whoever? No, because he knows that if they perform like that across 38 games they will win more than they lose.

What worries me is that Val couldn't see the pattern developing. We were out performing teams on xG through volume not quality. Not doing anything to rectify that cost him his job, rightly so.

There surely were other strikers available at the time? Some fecker picked him, and I'm convinced that Six had a big part in the choosing.

I get the point of the stats, but I don't need to be told that we dominated or not in a game, when I've watched it. As I know if we did or we didn't. If their keeper is pulling off loads of saves or their defenders are being run ragged and making heroic tackles and blocks then we dominated, even if we didn't get the shots away (which would lower the xG). I cannot think of many games that people told me that we dominated that I actually think we did, excluding the Blades / Cardiff stuff. Yet we probably had more games camped in their final third than we possibly ever have, we just never looked dangerous.

That's where I think a lot get annoyed by the overuse of xG. In the Pep vs Wolves/Palace scenario chances are that Man City really did batter them and the lesser club pulled off a Rorke's Drift, like we did against them... Again the xG for City might be through the roof, and in that case deservedly so, but again if you watched it you would know.

Maybe xG is more use for those who don't watch the games? As I said several times this season, Six played a style of football that you didn't need to watch, since most just went off the stats. It was akin to playing Football Manager, on full speed... oh we won... oh we drew... oh we lost... but in all cases the stats say we're brilliant, onto the next game...sort of thing, which is also probably where a lot of the boredom came from, since there is no fast forward button in real life.

Totally agree, Six must've seen the pattern, by watching the actual games, but he seemed to only care about what the stats said rather than what he saw. Which I do think played a major role in his downfall. The stats said we were amazing again...reality is our own fans were booing, that should tell you something.
Reply
#23
(02-03-2022, 09:15 AM)Derek Hardballs Wrote:
(02-03-2022, 08:57 AM)Ted Maul Wrote: Guardiola, Klopp, Mourinho, Bielsa, Tuchel, Nagelsmann all wrong. Stats are pointless. Val's inflexibility and his use of statistics are completely separate, although he does have a small point. I can think of three Hugill sitters off the top of my head that would have us 6 points better off.

Every manager with statistics available to them uses them to recruit players, pick teams, develop their tactical ideas and plans for the opponent. Especially for a club like ours where we can't just throw money at the biggest name players there are knowing their obvious quality, we have to rely heavily on statistics to see who is performing the best week in, week out. It would be absolutely negligent of us not to be doing this.

Take the example of a manager trying to sign a player. You can watch a player for hours and form an opinion, but it's only once you look into the stats do you know if said player will be an appropriate fit for your team and your tactical principles.

I didn’t dismiss stats, I criticised the over reliance on them by many, particularly XG stats.

The rest of your post makes exactly my point, rather than believe what your watching, using your own judgment you want to use stats to drive the decision making on signing players and who plays. The overriding stat that should govern any football manager are results. If a team is consistently getting positive results over a sustained period of time but the stats don’t back that up, do you change the winning team? Do you employ a player with a really bad attitude on the strength of his stats being good or do you take a more rounded and nuanced view of the situation and use your own judgment. 

Stats are one tool, they don’t tell the whole story and aren’t the absolute truth (fact), they will help a manager, but to rely on them as Val seems to have done is folly.

Agree that Val didn't spot the trend (all of our games were basically the same) and did nothing to change this. Lots of low quality chances and because we out performed the opponent he was satisfied, it rightly cost him his job.

Your point around believing what you're watching absolutely is a dismissal of statistics. If you think the very best managers sign players without looking at the key metrics for their position and the tactics that manager employs then you're completely wrong.

The example of the player with the bad attitude probably wouldn't get past the initial report.

(02-03-2022, 09:22 AM)MassDebater Wrote: There surely were other strikers available at the time? Some fecker picked him, and I'm convinced that Six had a big part in the choosing.

I get the point of the stats, but I don't need to be told that we dominated or not in a game, when I've watched it. As I know if we did or we didn't. If their keeper is pulling off loads of saves or their defenders are being run ragged and making heroic tackles and blocks then we dominated, even if we didn't get the shots away (which would lower the xG). I cannot think of many games that people told me that we dominated that I actually think we did, excluding the Blades / Cardiff stuff. Yet we probably had more games camped in their final third than we possibly ever have, we just never looked dangerous.

That's where I think a lot get annoyed by the overuse of xG. In the Pep vs Wolves/Palace scenario chances are that Man City really did batter them and the lesser club pulled off a Rorke's Drift, like we did against them... Again the xG for City might be through the roof, and in that case deservedly so, but again if you watched it you would know.

Maybe xG is more use for those who don't watch the games? As I said several times this season, Six played a style of football that you didn't need to watch, since most just went off the stats. It was akin to playing Football Manager, on full speed... oh we won... oh we drew... oh we lost... but in all cases the stats say we're brilliant, onto the next game...sort of thing, which is also probably where a lot of the boredom came from, since there is no fast forward button in real life.

Totally agree, Six must've seen the pattern, by watching the actual games, but he seemed to only care about what the stats said rather than what he saw. Which I do think played a major role in his downfall. The stats said we were amazing again...reality is our own fans were booing, that should tell you something.

It's not xG and statistics people should have a problem with, it's Val's poor usage and interpretation of them.
Reply
#24
How are the stats are created? Who / what judges what a good, half chance is? If it’s a person then that is subjective rather than fact.
Reply
#25
(02-03-2022, 09:29 AM)Derek Hardballs Wrote: How are the stats are created? Who / what judges what a good, half chance is? If it’s a person then that is subjective rather than fact.

xG specifically is calculated from a dataset based on thousands of shots previously taken that takes into account distance from goal, position on the pitch, position of defenders, speed of pass received, body angle of ball striker. As data models go it's pretty robust.

Stats are everywhere in the game. You can watch 90 minutes and still have no idea how a player would perform in another team or system. That's why shortlists are drawn up based off a manager asking his recruitment teams to find him players that meet; number of dribbles/90, ball recoveries, progressive passes rather than just saying they need a midfielder and getting a Sky Sports subscription.

I suggest you turn up to an FA coaching course and tell them you don't need stats, I've got eyes.
Reply
#26
(02-03-2022, 09:29 AM)Derek Hardballs Wrote: How are the stats are created? Who / what judges what a good, half chance is? If it’s a person then that is subjective rather than fact.

Sorry, do you think someone sitting in a room is deciding what xG each shot is?
Reply
#27
(02-03-2022, 09:35 AM)Ted Maul Wrote:
(02-03-2022, 09:29 AM)Derek Hardballs Wrote: How are the stats are created? Who / what judges what a good, half chance is? If it’s a person then that is subjective rather than fact.

xG specifically is calculated from a dataset based on thousands of shots previously taken that takes into account distance from goal, position on the pitch, position of defenders, speed of pass received, body angle of ball striker. As data models go it's pretty robust.

Stats are everywhere in the game. You can watch 90 minutes and still have no idea how a player would perform in another team or system. That's why shortlists are drawn up based off a manager asking his recruitment teams to find him players that meet; number of dribbles/90, ball recoveries, progressive passes rather than just saying they need a midfielder and getting a Sky Sports subscription.

I suggest you turn up to an FA coaching course and tell them you don't need stats, I've got eyes.

Derek (and I) aren't saying you don't need stats. Just that they shouldn't be the basis for all of your footballing decisions. Nor should they be used to cover up doing the exact same thing against every single team. Imagine Bielsa's spies watching Six, 'it's OK Boss, he's doing exactly what he's done all season'. Heck it became a running joke on here.

(02-03-2022, 09:22 AM)Ted Maul Wrote: It's not xG and statistics people should have a problem with, it's Val's poor usage and interpretation of them.

I think most of us agree with that line! Stats being used as an excuse to cover up poor performances would be the other.
Reply
#28
(02-03-2022, 09:11 AM)Brentbaggie Wrote: Is there a stat for the amount of unremitting boredom in a game?

yea, it's known as the TP factor.
Reply
#29
(02-03-2022, 09:39 AM)MassDebater Wrote: Derek (and I) aren't saying you don't need stats. Just that they shouldn't be the basis for all of your footballing decisions. Nor should they be used to cover up doing the exact same thing against every single team. Imagine Bielsa's spies watching Six, 'it's OK Boss, he's doing exactly what he's done all season'. Heck it became a running joke on here.

That's it though, in a game of such fine margins with big money involved you want the edge and that's what stats give you.

Val would have not only had our overall xG but also the xG of each individual shot, final third entries, touches in opposition box and he's looked at all those numbers and still done nothing to change it, therefore deservedly got the sack.

I keep going back to it but player recruitment is the perfect example of how stats are driving everything. You wont even be considered if your numbers do not meet the metrics that the manager in question is looking for.
Reply
#30
(02-03-2022, 09:45 AM)Ted Maul Wrote:
(02-03-2022, 09:39 AM)MassDebater Wrote: Derek (and I) aren't saying you don't need stats. Just that they shouldn't be the basis for all of your footballing decisions. Nor should they be used to cover up doing the exact same thing against every single team. Imagine Bielsa's spies watching Six, 'it's OK Boss, he's doing exactly what he's done all season'. Heck it became a running joke on here.

That's it though, in a game of such fine margins with big money involved you want the edge and that's what stats give you.

Val would have not only had our overall xG but also the xG of each individual shot, final third entries, touches in opposition box and he's looked at all those numbers and still done nothing to change it, therefore deservedly got the sack.

I keep going back to it but player recruitment is the perfect example of how stats are driving everything. You wont even be considered if your numbers do not meet the metrics that the manager in question is looking for.

And again, anyone who watched the games would also know that things needed changing. Nobody watching most of our games this season relied on stats to know that we needed to switch things about. The stats only got trotted out after the game. Check out most Match Freds to see even those who defended Six, calling for him to change things.

In what World did Hugill's stats get him a job, unless it was purely how far he runs in games?! Seriously. What stats does he have that mean he should be playing anywhere? Oliver Burke is another prime example. Lots of mediocre players keep getting massive deals, but their stats surely would not warrant it. Then again that's why I write code and am not a massively rich football manager. It's the one job where you can be pretty shit and still get more highly paid work.
Reply


Forum Jump:


Users browsing this thread: