It reminds me of the tongue in cheek joke from years ago:
“I’m level 55 gold cross prestige in Call of Duty. Of course the bouncer let me in!”
Matheus Cunha £62.5m release clause | Not a striker
- Thread starter dinostar77
- Start date
-
Nik (@Rood) and Imran (@Annihilate Now!) discuss United's fantastic win away against Athletic Club in the semi-final of the Europa League.
They also pick out the positives of United's 4-3 loss away at Brentford, and Nik reviews a day out at Old Trafford watching Manchester United Women qualify for the Champions League.
It reminds me of the tongue in cheek joke from years ago:
“I’m level 55 gold cross prestige in Call of Duty. Of course the bouncer let me in!”
Being a data analyst should ideally make someone familiar with basic statistical concepts like variance, but either someone is lying or that's not universally true.
I'm figuratively a data analyst and I can assure you all that this guy scored some very nice goals.
On the contrary, football clubs have been poaching data analysts from finance and other sporting areas for a while now. United just took a data engineer from Mercedes for example.
Thank you for the detailed reply.
Seems like the exact same on-field attitude that drove United fans to want to sell Rashford
Not according to his stats. Seems like this has been blown out of proportion.
Rashford comes across as the passive aggressive type. The type who wouldn't run and wouldn't speak up either, you just don't know with him. Cunha seem like the kind of guy the manager will know exactly what the issue is the moment he isn't happy, there will be disagreement but if you can get him in your corner, he'll fight but for you.
Think fans would have forgiven Rashford a bit more if, when in attack, he didn’t continually just lose the ball trying to take on defenders. He was like the worst of both worlds
Except Rashford forgot the bit about exploding with brilliance
Yes and they'll get no value from them until they genuinely understand the relationship between relevant data points and performance.
I'm probably missing something but you described data analysis.
Like his attitude with a bit of niggle, no harm in a bit of that rubbing off on our talented pack of young players.
Yeah I hate to sound patronising but JPRouve is right here, are you sure you know what data analysis is?
Yes - a lot more than you might realise.
A random data analyst that has no football knowledge won't even know what xg is unless is explained to them.
So the point stands that being a data analyst in itself doesn't mean one understands the implications of football specific stats better. I'm not saying that data analysis is not important in football. It's very vital, and if that is extended to data science, brilliant.
"Saying I'm a data analyst" doesn't suddenly mean you know what you're on about if you don't have an understanding of the subject area or what the metrics consist of.
The funny thing is, you don't need to know what xG is to know it correlates extremely strongly with goals scored. You could hand an analyst who has never watched a football match in his life a dataset and explain to him the simple concept of a football match and he'd be able to tell you players with higher xG are more likely to score goals than those with lower xG, he'd be able to tell you the variance in goals to xG and if a player was likely to keep scoring at that rate or not. That would be enough to tell you that Darwin Nunez isn't a good model for your expectations of Viktor Gyokeres despite coming from the same league (to steal a discussion point from the other thread), and that's without ever watching a match.
There's entire fields of data analysis which focus on blind analysis because it removes implicit biases which might lead you to overlook certain avenues of improvement. Data analysis at it's core is about identifying relationships, you shouldn't be relying on a deep technical knowledge of the subject because you're not a scout.
- Joined
- Sep 1, 2015
- Messages
- 20,042
- Supports
️
The bold is arguably one of the most important points of data science and recruiting expert data scientists as well as footballing people.
I think people overestimate how hard it is for someone with that background to understand data in contexts.
Figurative data analysts are the engine room of any successful team in the modern era.
15 league goals and zero penalty. Impressive number.
In addition to the goals, he offers a ball carrier and creativity. The fees might be high but I think it’s worth to take a risk for a player who has proved himself in PL.
In addition to the goals, he offers a ball carrier and creativity. The fees might be high but I think it’s worth to take a risk for a player who has proved himself in PL.
I'm only someone that has an interest in the field so my pov is likely wrong to some extent and you are welcome to correct me but there is a distinction between the role of a data scientist and the role of a data analyst which seems to be sometimes missed by people. A data analyst is going to collect, arrange and interpret historical data in order to spot trends or lack of trends when a data scientists is going to build predictive models based on historical data. As you pointed out data analysis, within a particular context, will often focus on blind analysis because you don't want to pollute the results while a data scientists may(will) try to tailor models to specific needs which will require a deeper understanding of particular fields and will introduce a certain degree of bias.
One few things I do know about xG.
- Its done to death on here
- It’s boring now
- People act like it’s the only possible way you can know if a striker is any good
- Helps people try to make themselves sound smarter
- Oh and it’s BORING
* Not an attack on you Dion, just my nemesis xG.
They would also without a doubt tell you that there are always outliers, and statistics are used to gain more understanding and awareness of important metrics, not accurately predict all individual cases.
Kind of, but not really. There are plenty of cases where analytics can absolutely rule out individual cases where the variance clearly takes them out of what would be consistent outliers.
Pretty much, although obviously the lines between both are generally very blurred in practice.
My only objection to any of this is the idea that it's valid up moan about what anyone wants to talk about. Statistics seem to be an area people feel empowered to tell others to stop talking about because they don't enjoy it. If we all took that approach to every topic I'd spend the entire day complaining about all the boring, useless shit you lot talk about
It's the kind of signing that would bring positivity for next season, it's something we definitely need.
Bit worried with some reports that he's a problematic character. Think we should always aspire to sign best players not only in terms of quality but mentality as well. It's what Fergie and David Gill always looked at I believe.
Somebody posted on here that he is very popular in the dressing room and around the club, very well liked.
They can't rule them out, because the return to the mean hasn't happened yet. They can with a certain degree of certainty predict it, but it's never guaranteed, that's the whole point of predictive analytics.
You can rule them out precisely because you can predict it to a degree of certainty. That's the entire point.
If someone has a regular pattern and then has one season which is significantly beyond anything anyone has ever sustained for more than one season, the degree of certainty you can say that it won't continue is functionally so high it's a guarantee. That's one of the foundational points of statistics.
So do we think he will be a good signing or is this thread just about data?
What’s Xg?
Computer says no.
Good thing I still use my trusty magic 8 ball, which says *shakes magic 8 ball*... hm, "ask me again later".
Last edited:
I don't think anyone believes that Cunha is gonna come in and score 15 league goals straight away because he did so this season do they? If they are then £62.5m is one hell of a bargain in this economy.
There's probably a clever-sounding way of working out the expected over-performance of high-variance statistical events and topping it off with Cunha's expected goals to come up with a number you'd want him to target for us next season. But the Data Analysts have announced themselves and I don't think I'll get away with whatever mistakes I make to come out with a number. So I'll just say that ten feel right.
I think if Cunha scores ten league goals for us next season he's probably been a success.
There's probably a clever-sounding way of working out the expected over-performance of high-variance statistical events and topping it off with Cunha's expected goals to come up with a number you'd want him to target for us next season. But the Data Analysts have announced themselves and I don't think I'll get away with whatever mistakes I make to come out with a number. So I'll just say that ten feel right.
I think if Cunha scores ten league goals for us next season he's probably been a success.
Born a day after our 99 CL win.
Written in the stars.
Written in the stars.
Especially if 7 of them are at Anfield
He'd be less likely to be a success (given that's 3 in the other 37) but more likely to be a legend. Okay, let's do it.Especially if 7 of them are at Anfield
- Joined
- Sep 1, 2015
- Messages
- 20,042
- Supports
- ️
I look at it this way.
For £92.5m (Cunha and Delap’s combined fees) do you think we’ll get above 20 goals?
I’d say yes.
Do we think we could get 30 from both? I’d also say yes but less likely.
What would a 20-30 goal a season player cost? I’d say up to £100m maybe more in the premier league cos that’s what Isak would cost.
So it’s probably good business.
Thirty league goals between the two would be huge. That's double out current best pair, and while you can't just plug extra goals in you'd imagine the team on the whole still gets output from others as well.
And both should make us more fun to watch which is still important.