Expected goals ratings

I am not a big car person, hence the functional Hyundai i30 instead of a flashier, more expensive set of wheels. Seven letters here in Western Australia, unless you pay a lot more to get a nine-letter plate.

Thereā€™s a guy who parks his Jag on the carpark at bescot for home matches with the registration
WA 15 ALL
I want it!

1 Like

Iā€™m surprised S Jackson was inside the graph lines for 16 /17 seasonšŸ˜ƒ

Move to Western Australia and you can have WALSALL in red letters on a white background for $500. It is available.

2 Likes

A moral victory for Walsall - I think a draw would have been the fairest result - but an actual loss for the Saddlers. Experimental 3-6-1 chalked up it as a win for Walsall because its margin over Cambridge was greater than 0.33 xG. I think at least 0.5 xG should be the line, but everyone is entitled to their opinion.

Good luck to Cambridge continuing to defy the xG data. I think we saw how bloody ordinary Cambridge is on Saturday, and that its League Two position is very flattering. I do not expect Cambridge to go up.

Iā€™m not sure that Cambridgeā€™s position in the league does flatter them. They have a goal scorer who will get maybe thirty goals this season. This from evidently very few chances. Having an asset like that gives you a real boost. The whole purpose of the game is to put the ball into the goal and Cambridge have somebody in their team who is very good at that, ergo they win football matches. That isnā€™t flukey or unfair.

In golf you can hit all the fairways and greens in regulation you like but if you canā€™t putt you still wonā€™t get very good scores because the essence of the game is to get the ball into the hole. Cambridge have the equivalent of a permanently hot putter in their team and Saturday showed us why it doesnā€™t really matter how many chances you do and donā€™t create, it is the ability to convert that really counts.

4 Likes

Good post, PT, and your golf analogy is an excellent one. I do not dispute scoring is the most important aspect of football, and it would appear Paul Mullin is a better than average League Two striker. But I saw enough on Saturday, combined with the xG data from 30+ matches, to suggest Cambridgeā€™s fundamentals are not very good, and it is only a matter of time before its results revert to the mean. I will put it another way: would you be interested in securing the services of any Cambridge player or management team member, apart from Mullin, based on what you saw on Saturday? I would not. I certainly would not be interested in Cambridgeā€™s manager just because the Us are second placed; Unitedā€™s xG numbers suggest they have been lucky at both ends of the pitch this season.

1 Like

Problem with saying Cambridge are in a false position, is if weā€™re using XG as a more accurate rating of a teams performance then Iā€™d flip it around and ask if you think weā€™ve looked like a top half team this season? We are according to XG. So are Mansfield who I thought looked dreadful when we played them at home. Play offs according to XG.

Itā€™s a very limited analysis and it favours teams that play a certain way. There are more ways to win a football match than just keeping the ball endlessly and occasionally popping off a shot.

When Walsall was able to field Jules, Adebayo and Holden, yes, I would say the Saddlers looked like a top-half team. But League Two is dreadful, so it is not saying much. And now Adebayo has left and Holden is injured, Walsall is weaker than the team that earned most of its xG data this season.

Lucky or effective? It is a thin line.

The very best Walsall sides have had a very good ā€˜keeper and a very good centre forward. I donā€™t know if there is xg data available for season 1998/99 but part of the Graydon miracle was winning 1-0 away from home with the home side very often feeling a need to report a robbery. Exactly how we felt on Saturday.

A team is obviously more than two players so the effectiveness of a back line will be a ā€˜keeper supported by a decent defence and the efficiency of a centre forward will be supported by a good supply line.

The data could be read to say that Cambridge have a very effective defence and a very efficient attack. Two qualities that justifiably impact their league position. We on the other hand has a reasonably effective defence but a very very inefficient attack which also justifies our league position.

There is merit in studying the data and the examples you give - especially Brentford - give it a ton of credibility. I think the higher you go in the game, the finer the margins, the more merit it has. Finely tuned athletes who perform at a very high and consistent level can, as a team, definitely gain a few per cent in their performance. At our level though I think it can muddy the thinking of coaches. A very simple game suddenly becomes very complicated. Not so finely tuned and very inconsistent athletes who need to concentrate on core skills getting brain overload from coaches who have the badges and have the data but forget the basics.

We have definitely seen this at Walsall over the last couple of seasons. Clarke obsessing over Zak Jules ā€œtraining numbersā€. Whereas the thing that matters most at our level is whether the player has the physical attributes to tackle, head it, control it and kick it vaguely in the direction they intended. You then slot them into a team where their role within that team plays to their strengths.

Back to the golf analogy, the very best players have metronomic swings. This means that all things being equal they will, 95% of the time say hit a 5 iron 220 yards and straight. Once you hit that level of precision and consistency you can then play a very scientific game within the variables of different courses and conditions. For poorer players it is about getting to the point where it goes mostly straight and somewhere around 175 to 225 yards. To get better than that needs thousands of hours of practice so the best thing those players can do is just practice hitting the ball. Same in football. At the top of the game the first touch and delivery are more or less givens so it becomes a nuanced game of space and time. The winner often decided by football intelligence (of manager or player) rather than the ability within the feet. Whereas in the lower reaches of the professional game it is far more about getting better at the core elements of the game and not obsessing over stuff that assumes you can trap the ball, run with it and pass it.

1 Like

Touch and go at very best (and throwing Holden in is a tad disingenuous because the XG figure takes into account he hasnā€™t played those games) Iā€™d say but ok.

xG has its limitations, but it is not a very limited analysis. I have been monitoring it for years, and I think the only teams it really disadvantages are teams which set up to absorb heaps of pressure before hitting their opponents on the counter. Atletico Madrid is the example that leaps to mind for me. But with regards to most teams, I think xG is a very reasonable metric by which to assess their quality.

But thereā€™s tons of teams that play that way in this league? As you go down the pyramid fewer and fewer teams keep the ball immaculately and look to play aggressive football because they play to their strengths. Which are normally physical.

And I can think of even more examples than Atheltico. Wolves have won the championship and got into Europe playing that way. Jose Mourinho has made a career out of it. That is why itā€™s limited.

Not to the extreme of Atletico Madrid under Diego Simeone or, say, Leicester under Claudio Ranieri. It is only the extreme cases with which I have an issue, and there are very few of them. And I do not see what physicality has to do with anything. Possession and territory, yes. Physicality, no. xG simply attempts to put a number on the quality of chances that teams create and concede. Physicality has nothing to do with xG numbers. Zero. Zip. Zilch. Nada.

I am not advocating using raw xG numbers - and by that I mean the match charts and league tables - for anything more than data-driven snapshots of matches and teams relative to results and positions, respectively. Football is a relatively simple game, particularly at League Two level, and the basics are what are important on the training pitch.

Letā€™s pretend for one moment I did not watch Cambridge versus Walsall. The xG match chart suggests it was a game of very few chances that Walsall edged on the balance of play, but Cambridge stole with a worldy goal in injury time. That would be a pretty accurate account of the match, wouldnā€™t it?

1 Like

But there arenā€™t? Iā€™ve just given you examples off the top of my head. Most of our division plays that way.

Physicality has to do with playing to your strengths. In this league that is very rarely tippy tappy XG friendly football.

I disagree with your premise. That is OK, isnā€™t it?

1 Like

What premise? I donā€™t know what you are talking about? I gave you those examples. Disagreeing is ok but Iā€™m not sure what you are disagreeing with? The fact most teams in this league sit back and look to be solid first and foremost?

I do not agree with your examples. You are entitled to your opinion. I am entitled to mine. They are opinions, not facts.

1 Like

Itā€™s not matter of disagreeing itā€™s fact, Iā€™m afraid. Teams in this division play that way and so did the examples I gave. You can disagree it matters, but the fact teams play that way is absolute fact :man_shrugging:

Too many people confuse opinion and fact. You are indeed entitled to your opinion, your not entitled to your own facts Iā€™m afraid.