Holebas - 4th best full back in the Premier League

http://talksport.com/football/top-2...df8e6fbee-RSS_EMAIL_CAMPAIGN&utm_medium=email

And 20th best in Europe.

 

Agree
Voted him player of the year!

 

And the most bonkers full back in the Premier League

 

Which makes him one of the best (to watch) !

 

This can't be right. I remember reading on the forum that he's a ***** useless stinking ****** with a ****ish attitude who can't defend for toffee.

Yeah. He was my player of the season too, but still.

People will be saying that that lazy **** Capoue ran eighth furthest in the PL next.

PS Looking forward to the 'Amrabat beat the third most defenders' statistic coming out in the next week.

 

Amrabat beat that third choice defender against Hull, but that's about it

 

What did Amarabat beat the defender with?

Is that why he was regularly penalised?

 

I cannot disagree with this, and it only goes to show just how wrong many people on here have been about him. I know he may not be "Mr Likeable", but his committment and passion, to what afterall is his job, cannot be questioned IMO

 

Ridiculous article. He had a decent season but he's clearly not the 4th best full back in the league.

Before people start worshipping WhoScored ratings as the ultimate arbiter of performance, consider how they've rated some of our squad this season:

Prodl was only the 13th best player in our squad
Guedioura had a better season for us than Doucoure
Our top three players of the season in order were Holebas, Kaboul and Capoue

https://www.whoscored.com/Teams/27/Show/England-Watford

 

To be quite honest, after years spent watching all of us rate our own squad I'm far more inclined to believe algorithms than our own assessments. We're a notoriously fickle bunch and literally every player in the squad is good or carp depending on who you might ask.

Emotional attachment tends to muddy the waters significantly.

 

Really? The algorithms think there were 12 players in our squad who were better than Prodl this year! And that Guedioura performed better for us than Doucoure. And that Kaboul and Capoue were our 2nd and 3rd best players.

I don't think a single fan would agree with any of those assessments, and they'd be right not to. Football is a dynamic game where performance can't be evaluated quantitatively alone, there has to be a subjective qualitative aspect too. These kind of algorithms work brilliantly in stop/start games such as cricket and baseball but they are seriously flawed for flowing games like football or rugby because they tell you nothing about the quality of the actions taken, just the quantity.

 

You seem to be implying that you disagree with Arakel, yet you go on to nail his point exactly.

Anything that is subjective, can only be viewed objectively as an opinion. The stats may be flawed, but they don't express an opinion.

Gueds, Kaboul and Capoue - the unfashionable, the miss-fit and the miss-understood, have been easy targets this season, and once one person on the forum slags a player off, there are many that love to join in with all enthusiasm, and after a while it becomes the perceived wisdom. But in the games Gueds played, no matter what the perception of his quality, we looked a more likely team than when he was taken off (I don't expect any of you to remember that, or even to have noticed it, seeing as it wasn't mentioned on Match of the Day), and despite this forum, I know I'm far from alone in thinking that. Not saying he was the greatest player, just open to the possibility that he may actually play a positive roll in a game and be recognised for it; many on the forum were discussing how he wasn't good enough before he'd even got his shirt on, so there was little or no chance you'd see anything decent that he did, so no wonder you are surprised when someone say he actually did OK. Kaboul put in some great performances, but unfortunately, on the forum, it was the worst of his performances that were remembered, so you question the ratings he received. And Capoue is judged not on what he does, but on how he looks as he does it, employing a languid style, not completely unlike Stewart Scullion, as he fails to break into the top seven elite of ground most covered.

It's OK. People are free to have their opinion. But it is gratifying to see the confusion and indignation when those opinions get challenged by others, or worse,by the facts, despite some of the views expressed in this rarefied environment.

 

They aren't the facts, they are some facts. It seems to me that the stats are very basic, usually binary events, and they favour positive events over negative. I suspect if they correctly accounted for things like losing possession in a dangerous area, not making the correct pass that was available, or failing to mark properly, we would see a very different set of rankings. I'm not claiming those rankings would necessarily align with fans' views.

 

That is a very fair point. But now you are getting into IF statements, the application of which, in the type of conversation we have on the forum, tends to invite even more subjectivity, expressions of anxiety as opinion, and parallel universes that never existed presented as fact.

Until someone produces those other facts you 'if', what we have are the facts.

 

Yes, I get it. You think of yourself as "the thinking man's Watford fan", taking contrary positions to show how you're not part of the crowd and you have a better understanding of the game than the rest of us, who obviously wait around for MOTD and then just repeat what they say (notwithstanding that a lot of the post game analysis happens on here before MOTD is even shown, and that Watford are very rarely analysed in any detail at all). Good for you.

However, your whole ramble above depends on whether you're prepared to answer three questions in simple, one word answers. No verbose, patronising waffle just a simple Yes or No:

1. Was our Player of the Season Prodl, in fact only our 13th best player? Yes or no
2. Was Guedioura a better performer than Doucoure this season? Yes or no
3. Were our best three players last season Holebas, Kaboul and Capoue in that order? Yes or no.

If you answer No to any of those, then your whole argument is nonsense, as you also don't agree with the WhoScored performance ratings. Incidentally they're partly subjective too as they rely on humans interpreting all of the actions on the pitch (e.g. who won a tackle, whether an error led to a goal), as well as those humans correctly identifying the players involved each and every time.

Anyway, look forward to a succinct response with three Yes or No answers. I won't hold my breath.

 

The final decider if whether a player is good or not IMO is @reids scouting notes

 

Some of the statistics that Whoscored use to rank players and some of the weighting that it gives to certain actions are really quite ridiculous. Their stats frequently produce silly results like this, and they frequently clash with general fans' view of a player's contribution to their club. I'm not arguing that Holebas was bad, he was fine and one of our more consistent performers, but attempting to make any kind of argument that Whoscored provide any definitive analysis is wrong.

As JW said, they aren't 'the' facts, they are some facts, and they are very much open to collective questioning.

 

Steve, yet again, you prove the point that is being made better than I could possibly do. And thank you for attempting to insult me by accusing me of thinking before posting.

I have to ask you. After everything in this thread, what has my opinion, or anyone's opinion, got to do with the facts? Do you not understand? That is at the crux of my argument

I have, below, given you one word answers, but I have then explained the answers so they can be taken in context, otherwise they are pointless.

1, No. Prodl was POS, and well deserved (not mine though, I went for Holebas); but POS isn't decided on WhoScored data is it? Did I say it was? POS is entirely subjective in its nature, and WhoScored's factual data does not cover all aspects of a player's game, just what they achieved in certain areas.

2, Yes, according to WhoScored, based on their data collection. End of. Silly question seeing as you already knew the answer. Is he a better player? That is another question, which requires a far more subjective answer. Is Gueds 5hite? I don't think so, and I think that what he actually achieved in a game should be taken into account when judging a performance, rather than the more simplistic 'he's 5hite so he must have been 5hite'.

3 No. Silly question. WhoScored is not a measure of who is best. It is a measure of certain elements of their performance.

My point about the facts portrayed in WhoScored is not that they tell you who is best, but that they show people who have a simplistic, emotional view, based not on whether a player has performed well, but on whether they like them or not, that actually, factually, and objectively, that player achieved more than they were given credit for.

Doesn't mean they were the best, just means that they were better than you were prepared to accept.

Hopefully they'll explain that on Match of the Day for you. Baahh.

 

I don't see how any measure that had Gueds better than Doucoure can have much credibility.

 

It doesn't say Gueds is a better player. It says he was rated higher based on the parameters measured by WhoScored.

It suggests, regardless of perceived ability, that in the games he played, Gueds contributed more to the outcome of the game than Doucoure did in his appearances. Hardly a surprise when you consider the games they appeared in.

 

All this shows is that the parameters are flawed as a tool for evaluating performance.

 

So in summary, the whoscored facts are facts and therefore cannot be disputed. However they give no indication of whether a player has played well or is fact any good at all. They come up with conclusions that very few of our fans would recognise and can probably be completely disregarded and no point using them to get a clue about potential signings ?

 

On the topic of stats, I do think it is inevitable that machine scouting will become better than human scouting. I think the main thing that is missing is for machines to interpret 'good' and 'bad' decisions from a player.

If I was Gino I would start up a Watford Research private lab with people from Google and try to develop a scouting system based on deep learning. 10 million pounds is pretty small change for a player, but massive in terms of a research grant.

 

The WhoScored facts, are facts. Their validity can be queried, they are not perfect, and even I don't say they are the mark of how good a player is, just that they are a valid empirical measurement that may be used to inform an opinion, or, more to the point on this forum, be used to consider the validity of an opinion. That is, 'Gueds is 5hite', 'but the facts indicate he was effective', 'but, but, everyone says he's 5hite, so the facts must be wrong', 'Oooookaaaaaaayyyyyy'.

They do not come up with conclusions. They provide you with the information on the players performance. You/We make the conclusions. Data/facts are not an opinion. This is another 'follow the sheep in front' moment, where some on here believe they need to be told what to think. WhoScored and Opta are there to inform your opinion, not to provide you with an opinion. Please understand that.

When it comes to potential signings, fan opinion is the last thing you might want to go by. Nyom or Gomes being a prime examples; when put beside that, WhoScored suddenly seems a far more reliable source of information.

 

He's my hero.

 

Haha, I like Holebas but he definitely isn't the 4th best in the division.

 

Algorithms can be flawed, yes. But so are the assessments of fans, who are largely led by emotional conclusions and outbursts, forced to make assessments on incidents from memory (which could be faulty), subject to equating recent form with form over a whole season, and prone to group think and binary assessments. Much is made in the legal arena of the unreliability of eyewitness accounts. Individual fan assessments are, in essence, exactly as reliable.

Stats certainly have their own limitations. For example, statistical comparisons aren't useful for comparing central defenders to attackers. Where they are useful is in comparing central defenders to central defenders, or left backs to left backs. Even then you have to be careful, since an ultra-defensive team and a gung-ho-overloading team will see the roles of some players vary significantly, even if their teams play in the same formation.

There are definitely outliers where a player can game the system and flatter to deceive statistically, but they tend to be obvious. Heck, I'd argue Ighalo was a perfect example of this: banging in the goals and achieving a huge score on paper, but not really contributing a whole lot to the team to the point that he was unceremoniously shipped out once his purple patch ended.

Holebas is a player who was rated by many of our fans, and slated hard by others as someone we need to ship out. The statistics back up those who think he did well for us by comparing him favourably in to other players in the same position. Many fans dislike Holebas due to his perceived attitude problems, and colour his performances based on that. It's not different to how many fans have blindly defended Deeney's performances during some of his form slumps because they're emotionally influenced by what Deeney has done in the past.

Furthermore, lets not pretend individual fan assessments are anything more than shoddy internal versions of sites like Whoscored, limited by personal biases and individual recollections of what happened (and for many, recollections of experiences under the influence of a lot of alcohol). Implying that those rating players for third party sites are somehow less capable of judging a won aerial duel versus a pissed up fan in a stadium with a bad view is a bit silly. Organisations like OPTA do this for a living. They've honed their craft over time and it's well understood now. Fans assess players in basically the same way as OPTA, we're just far less scientific with how we go about it.

Stat crunching is certainly not the first and final word, but I place far more weight on them when used correctly than I would on the opinions of individual posters subject to echo chambers. There have been many examples of prevailing opinion differences between fans on this site, the Wobby and Twitter to illustrate why.

 

Perhaps not, but I thought, over the last two seasons, that he is one of the players we have who could play anywhere in the PL. Not saying they're gonna buy him, but IMHO, he could play for any team without looking out of his depth, and possibly the best player we have with the ball at his feet.

 

Holebas needs his own tv show

 

And on that basis. Thread closed.

 

If you have not watched it, the film Moneyball is a true life film about a Baseball team who chose to use computer stats to build a cheap team to successfully compete against the rich teams.

 

Holebas is underated by almost everybody. My POS - very few mistakes throughout the season, certainly not directly from him anyway. Played the last game mostly injured and did his best in ludicrous circumstances.

 

Algorithms utilise parameters the framing of which is based on subjectivity. Facts are effectively conclusions drawn from a body of data that is, at various times, more or less complete. When more germane data becomes included, the initial conclusions may need to be significantly amended, resulting in the specific "fact" involved becoming different.
As the old adage goes "there are lies, damned lies, and statistics".

 

Brentfords owner uses the same technique with both his teams (Brentford and Midjtylland). Brentford have one of (if not the lowest) wage budget in the Championship and Midjtylland won the Danish title for the first time in their history by using such a system.

 

So, at what point do you make a decision that a fact is what it is purported to be?

Or do you simply not trust any factual data? As you seem to imply none can be trusted?