Never thought I'd see an article about Arsenal here.
The "secret" purchase has been talked about for quite some time now. We'll see if it pays off. In the mean time, while we wait, Manchester United has spent a king's ransom on a teenager and so far it's looking like a great bet (yes, I know it's early still). I'm a strong believer in analytics in basketball and baseball but I wonder if in some sports, like soccer, it is harder or perhaps not possible to quantify a player's effectiveness in a scientific way. I know my skepticism sounds just like what was mentioned in the article.
Having said all of that, Arsene revolutionized scouting when he came to the PL by bringing on somewhat lesser known names from across Europe for prices that proved to be bargains. Some of those players turned into legends. Maybe with a little help he can reinvent scouting again.
I think it's fair to say Arsene's ability to buy bargain players has died up. Gone are the days of Henry/Viera/Van Percy coming in as mediocre players only to be indispensable a season later.
I think if you look at some of the more recent meteoric rises, they all seems to come from lesser leagues or clubs, BVB/Juventus/Atlético Madrid/Porto, and to some extent Liverpool/Spurs (Bale/Suarez/Sterling/Coutinho/Kane).
I think signing a player with potential is only part of the issue, the key is playing, developing and nurturing them. With teams like BVB/Juventus/Atlético Madrid/Porto they are either in easier leagues where a player doesn't have to be the finished article to excel, or financially constrained to not be able to purchase the finished article. With Spurs and Liverpool they fall into the later category so are forced to give youth players a chance.
Take Kevin De Bruyne as an example, an obvious talent purchased by Chelsea, but not played because he wasn't quite finished (perhaps only in reputation rather than ability). Sold to play in a lesser league where he excelled, and then purchased for 50 million where his stature and starting place are then almost granted.
These days I think Arsene is stuck in a position where he has the money to spend, but ideally want's to return to his roots of nurturing young players. However the fans and board aren't willing to gamble on youth, in favor of signing 40/50 million finished articles.
A few million on an analytic company isn't going to break the bank, and may be more valuable in opposition analysis than finding the next rough diamond.
I think the analytical approach is still working for Arsenal; Monreal, Koscielny, Coquelin, Bellerin, Carzola, Giroud - non of those guys were marquee signings who cost loads, but all are now indispensable and could arguably walk into most teams in Europe.
The analytical approach is clearly working for other teams too. The game is slowly changing and people are finally seeing the worth in the stats.
I would argue they are typical squad players for a team of Arsenal's level. I would be very surprised to see any of them get poached from another champions league club, or have a statue erected outside the Emirates.
Analytic no doubt played a part in their signing, but it's hard to argue that they are exceptional examples of recruitment in the way Lewandowski or Kagawa were to Dortmund.
Wow - Had to mention Harry Kane on here huh? He is a real star, did you see his wunder(own)goal the other day? Took his tally up to 2 goals on the year! Our carl jenkinson has more than him! (ef the spuds).
In all seriousness, I don't think arsene's ability to unearth hidden gems and turn them into 10/20/30 million $ players has completely gone. I mean FFS hector bellerin (not really unearthed from some hidden continent in Africa, he was bought as teenager from Barcalona) was a complete unknown and now is dead locked for the first team spot for Arsenal. This is the unknown youngster who, within a span of 1.5 years, went from nothin to first team start. He just played against Bayern Munich which is possibly the best team in the world, was able to keep Douglas Costa off the score sheet (top top winger/stricker) and to top it off, made the most insane interception I have ever seen. His pace is literally electric and even sitting at home on my couch I can feel the people in the Emirates standing on edge when he starts racing forwards.
> not really unearthed from some hidden continent in Africa, he was bought as teenager from Barcalona
This is HN, can we have some geographical accuracy and keep the casual stereotyping to a minimum? First of all Africa is itself a continent, second of all it is not exactly unknown.
Just because some soccer-watching people don't know much about it doesn't mean that comments about "some hidden place in Africa" are particularly appropriate for a more general audience.
I think those players all have the potential to achieve legend status at Arsenal (maybe not Monreal so much, and maybe not Giroud though for the latter I'd argue because he's under-rated by many) if they a.) Keep going the way they're hoped to for a few years (improving for youngsters like Bellerin, staying the same quality for Cazorla) and b.) are instrumental in bringing a few more trophies to the club. It needs a mix of personal quality, longevity at the club and also team success.
That said, even if none of them go down in the books as legends, they can still be excellent examples of great value purchases.
I would argue they are typical, perhaps good signings for a club such as Arsenal. They were all however known, and already playing at big European clubs.
None were say playing for say Halmstads BK or unknown to everyone in the top tear other than vai some obscure stats system...
To your point, I'd say two of his best signings in recent memory were Cech and Özil -- both mature players who were already recognized as hugely talented on the international level.
However in fairness to Wegner, I think Danny Welbeck was identified to be a potential "hidden gem" but unfortunately he has been plauged by injuries. And it's hard to criticize one for being unable to find the likes of Henry.
I don't think you can class a 15 million pound Welbeck, who was playing for Manchester United and the England senior team as a "hidden gem" when compared against Shinji Kagawa who was playing for Cerezo Osaka in the J League and signed for €350,000...
Related question from a non-sports person, about those 'bargain' prices. How long does it take for a player to become legendary, and when does their price catch up? Is the contract a bargain because it locks-in the unknown player at a below-market price even after they become legendary? Or is it a bargain because the contract lasts a season or two and establishes the player's skill, after which they have inertia on the team and can be signed at a new market-rate contract with less bidding-wars that would be the case if they were from another team?
Semi-unknown players from mid-tier clubs simply cost far less in transfer fees than established high-profile players. A classic Arsene Wenger signing was Freddie Ljungberg (https://en.wikipedia.org/wiki/Fredrik_Ljungberg). He was signed from a team called Halmstads in the Swedish league. While Halmstads were doing pretty well in the Swedish league at this time, they were unknowns on the world scene. Since Arsenal had little competition in signing him and since Ljungberg was pretty unknown they were able to pay £3m - which is not a large sum at that level level. For some context, at the time the most expensive signing in the world was Denilson who Real Betis bought for £20m. However Ljungberg went on to play 216 games for Arsenal, scoring 72 goals which is a considerable number. For comparison Denilson - who was played in roughly the same position - played 186 games for Betis and scored 13 goals. Ljungberg is generally considered to be an Arsenal legend, while Denilson is thought of as an embarrassing waste of money and gets barely a footnote on the Real Betis wikipedia page.
As for what is considered a "legend" I really don't know, football is weird like that. It could be that you achieve this status by being an aggressive no-bullshit kinda guy (Vinnie Jones at Wimbledon), a world class player who stays at the same club his entire life (Ryan Giggs @ Man Utd) or just being an insanely prolific striker who scores for fun .
On this topic though, various newspapers\magazines\sites periodically publish "worst transfers" lists which are basically filler (everyone who cares already knows who will be on the lists), but which could be surprisingly enlightening for someone who's not familiar with football's transfer scene. This one's a bit advert-y but you'll get the gist: http://bleacherreport.com/articles/1610214-13-stupidest-tran...
In football (soccer), more so than in sports like baseball in America, players' costs are often more in the price tag when getting the player than their wages. A very good player might make £10m/yr in wages, and cost £50m to buy from a different team.
Unknown players will be cheaper on both, but even if you picked them up on a one year contract and they immediately proved that their next contract should be highly paid, you've still saved a potentially huge amount on buying them before they were worth a huge signing fee.
Gradually salary will catch up based on how well they play and based on the market, and the team that found this cheap player can either stop considering him a cheap player (but still be glad he came in cheaply), or can sell him on for a profit.
The top 10 salaries in the premier league range from ~£10m/year to ~£15m/year, so consider that buying a top player and keeping them for 2 years can often cost £20m-£80m. If you can buy a player for cheap and pay him not a huge amount, you can afford to spend a few years turning him into a star player and still come out way ahead even if in those few years he wasn't giving you value other than turning into a better player.
How long does it take for a player to become legendary? "Legendary" would normally be reserved either for the very best of the best, or on a team level based on both quality but also loyalty - as such you'd rarely see someone be called a legend before spending quite a few years at a club. But a player can go from relatively unknown to "shit he's really good" in as little as a season potentially - maybe he was playing in a league that doesn't get as much attention from countries with the bigger leagues, maybe he needed better team-mates around him to bring out his best, maybe a manager or coach helped him change his playing style very quickly, etc.
In football at the moment, there are two elements to price - wages and purchase price.
Clubs 'sell' players who are already contracted to them, to other clubs, in order to have the contract nullified. The price can be tens of millions of euro. The buying club will tie down the new player to a long contract as part of the purchase, perhaps 4 or 5 years.
Wages are low for young players, but will be increased rapidly if the player emerges as a real talent that other clubs would want, which gives the club the chance to extend the contract through renegotiation. A longer contract for a highly rated young player would mean higher wages also as part of the bargain, although contracts can be renegotiated and extended with higher wages at any time, if the club views the player as an asset that they wish to keep.
When they extend that contract it generally means that another club who tries to buy will have to pay more, and usually wages are at least matched and often increased by the buying club.
To expand on that a little with a software analogy, is it a bargain because they hired a talented intern who graduated into a full-time employee, or is it a bargain because they hired someone talented from a different country who has lower salary expectations?
Well, very unlikely you will pay some 30+ aged footballers (except GK) much higher value regardless of which countries they were from. But for youngsters, they are full of potential, so a tool to help the club to scout is very helpful. In the past, they are relying heavily on "scouting agents", and that's bit of an art than science with data backed.
Less about wages, more about signing on/purchase fee. The analogy doesn't really work because in the non football world employees are not bought and sold in the same way.
There is occasional talk about doing away with the system, as it goes against a lot of employee rights principles.
It's incredibly random in football when it comes to assigning value to players, at least in Europe.
Take the example of James Rodríguez[1], a young Colombian player. His "value" suddenly skyrocketed[2] when he scored this wonder goal[3] in the 2014 World cup. Next thing, Real Madrid snapped him up for €80million, the decision at least partially influenced by Real's "Galaticos"[4] transfer policy, and the fact that Real can increase their commercial influence in a Spanish speaking country because Rodríguez is Colombian.
Other factors that influence a players value:
- time left on contract (longer time means more money)
- "marketability" e.g. player's ability to sell lots of club merchandise, and improve the club's brand
- perceived desperation of the buying/selling club (do they have a desperate need for a top class striker to make up for a lack of goals?)
- if there is a bidding war between clubs (e.g. young French striker Antony Martial was subject to a bidding war between Manchester United and Chelsea, dramatically driving up his value.)
- perceived transfer war chest of the buying club
- player's age
- if the player came off the back of a successful international tournament, which dramatically (sometimes artificially) showcases their skill. Sometimes teams will buy a player ahead of the tournament to prevent a later bidding war.
- skill of the player's agent (e.g. Raheem Sterling to Man City)
- then players skill and so on...
Every deal is a virtual once-off, it's always a bargain in hindsight[5] (whether you pay little or big money or not) and the player exceeds whatever the initial expectations were.
I'm not sure Martial is as risky a bet as people make out, it's not like United don't have a similar analytics team in place (every premier league team does), doing the same kind of analysis.
Arsene certainly revolutionised scouting, but I think now it's a lot more well understood by other managers.
The problem with applying analytics to Martial is the limited dataset - he only made about 50 appearances in his career before joining United. It's looking a good bet so far, but it's definitely a punt.
With young players, you take quantitative factors more into account than an older player. Staff and players at Manchester United have mentioned in interviews how Martial is extremely mature for a 19 year old, he is married with a child and obviously wants to provide for his family. The fact that he is French with limited English, means the tabloid horseshit rolls off his back and doesn't detract from his game.
There's still lots of great football stats work being done out in the open, at http://statsbomb.com/ for example. One of the founders of StatsBomb, Ted Knutson, went on to be hired by Matthew Benham to do player analytics for his two clubs. There hasn't yet been the sudden land grab that there was in ice hockey where all the best analytics bloggers got hired and silenced pretty much overnight, but clubs are getting far better at this stuff.
Between Opta's human-based on-the-ball event tracking, and TRACAB's computer vision player tracking solutions that they provide to the Premier League, there's a bunch of interesting computer science and business going on around this stuff.
As an Arsenal fan, this somewhat surprises me considering Arsene Wenger seems to be a fairly old fashioned sort of coach these days.
It seems an eternity ago that he swept into the league with fresh new ideas. Nowadays many people (wrongly) consider him a stubborn old man who has been passed by by the modern game.
I disagree (and I am too an Arsenal fan). Wenger always brought in these new blood who are mostly below the age of 23, and I guess that's the trick he used to scout them.
I also think he is not the type of guy brag about his tricks.
But this is just for scouting. I'm yet to see someone can get all the data on each games, and help in coming up with tactics and strategy for planning the next match. Perhaps employ some ML too!
What? Couching is not all scouting... no one's denying that Wenger is always on the look out for young prospects.
But Wenger is a sort of 'old fashioned' coach. He prefers focusing on how his players can do better, rather than trying to focus on the weaknesses of the opponent, which is definitely not 'modern'. This has been slowly changing the past few seasons however.
That being said, I'm definitely no #WOB just in case you think that.
I do agree with your above points, but he has his football (sorry American) philosophy which not most fans understand.
The past decade, he cut his spending to just build a new stadium for Arsenal, budget is tight, so what to do? Bring in the new blood, and the hard data + analytics help!
I'm hoping to see he can also bring in some data-driven approach on his on-field tactics and off-field strategy.
If by football philosophy you mean "the thing about Arsenal is...", skc got it right I think because 15 years ago he was definitely considered 'progressive' or 'modern'. But now I think he's considered "a stubborn old man" because of a "lack of a plan B...". (Of course, past few seasons that "lack of plan B" criticism is not really relevant anymore, but you probably know what I mean.)
About football the sport, I've always seen him as 'progressive' because for example he has always advocated for video replay technology to be used, goal-line technology to be used etc. And so it's definitely possible that he's not afraid of using all the technological help he can for coaching.
It's just that he has rejected using 'technology' to help his coaching in the past. By which I'm referring specifically about him not using video replays of how his opponents play to coach his players with (which has since changed). But I think that is more the case of just not being part of his coaching philosophy, rather than not wanting to use 'technology' or something.
Gotta disagree there. Take the game against Bayern a couple of days ago - he knew that Guardiola prefers to play possession football so he let him do that. Arsenal defended deep in response, hitting Bayern on the counter and reaped a 2-0 result. He certainly exploited the flaws in Bayern's game.
There were no big flaws in Bayern's game that day. Arsenal sat deep specifically to shut down what muller and lewandowski could do, but I wouldn't call that exploiting. In fact Arsenal didn't really 'exploit' or target any aspect of Bayern's game (it's hard to find flaws against "top top class" teams anyway), they were just really well disciplined and played some really good high tempo counter attacks. The lay-low and counter attack was just a good gameplan, not really an 'exploit'.
But Bayern still mostly dominated and played really well. Arsenal's goal was mainly due a horror error by Neuer, he really should have caught or punched that. (Yes it's a set piece so anything can happen, but other than Walcott's header, Sanchez's chance and Giroud's late header from a corner, I don't really remember many more threatening set pieces, so Neuer should have really dealt with that, especially after the Walcott save.)
The only 'flaws' that Bayern had were that they didn't finish. Costa should have scored. Lewandowski should have scored maybe twice even though it was a quiet game for him by his standards. (Sanchez's mistake was what lead to Costa's chance on goal too.)
That last goal by Ozil was really just more of a case of Bellerin being a freak surprisingly right at the end of the match rather than Arsenal specifically exploiting anything. Arsenal was not high pressing (as a team) on that play, it was just a mistake (slow pass) combined with the incredible anticipation and speed from Bellerin (not to mention the desire to bolt into the box rather than jog to the corner flag).
In fact, you could clearly see throughout most of the match that Bayern was the one that kept wanting to target Nacho. Which was pretty weird because if they did their homework, they should have known how massively impressive Nacho has been recently.
> This has been slowly changing the past few seasons however.
And despite all of what I just said, I am acknowledging that Arsene isn't as 'old-fashioned' as he was maybe 5 years ago. Because that comment was more of an observation in general of the post-Highbury era, not literally the past few days when it's obvious that there's more to Arsene's game nowadays.
If you wanted to use a recent example anyway, the game against United would have been a much better example imo (league games is generally when you should be 'exploting' anyway, not in the cl, because you should be expecting anything in cl games, just because of the occasion). You could clearly see that Arsenal specifically targetted United's slow midfield by high pressing on them (Walcott's and Ozil's workrate was impressive). And they were basically one-twoing their way past Bastian "cannot run no more" Schweinsteiger the whole match.
This is an example of a good movement by an attacking player (Munir, player from Barcelona starting at the top) that helps create the goal and still won't be traditionally credited as anything (didn't score or assist or even touched the ball in this play): http://gfycat.com/TotalEmbellishedDesertpupfish
At some level soccer is about creating space for yourself or for your teammates, this is what this player does (following Suarez' clue) and hopefully the new data companies are capturing this (versus silly stats like possession).
The owner of SmartOdds betting company, Matthew Benham is applying his company's statistical insights to the clubs he owns. This has proven quite successful. Great to see Arsenal doing the same.
http://www.theguardian.com/football/2015/jul/27/how-fc-midtj...
It has to be said, the article gives remarkably little insight into what Benham's statistical system actually offers (and of course what was so objectionable about it to the successful manager who chose to part company with the club over it). I mean, any football club in the world has access to data on "any two-footed left-back in the world of a certain standard, aged 22-26, who has not been injured for the previous 18 months"; we had games that could do this pretty reliably in the 1990s.
Football seems like a particularly difficult sport to try the Moneyball approach with because it's largely a battle of team cohesion rather than individual skills, and even apparently novel and meaningful heuristics one might devise like "percentage of forward passes completed in opposition half over the course of a season" are likely to be heavily influenced by the tactical approach of the player's team. Even if your telemetry data is incredibly comprehensive, it's going to be difficult to teach a computer the difference between chances created because the timing of the pass was exquisite and chances succeeded because the covering defenders were out of position or made a glaring hash of the interception; not something you'll ever find coaches or commentators struggling to decide on.
As sports are now much more data driven, is there any other analytics are doing similar? In this case of Arsenal, they purchased it for "scouting" purpose. I would also love to see if some sport clubs can leverage the data and make sport management as a predictable science.
Opta certainly hope so. They have, and licence access to, a colossal amount of sports data and I imagine clubs that don't have the resources to generate/collect their own will probably use a service like theirs.
"StatDNA’s database is now Arsenal’s database, so if Wenger wants to sign a defender, who is, say, 6ft 4in, an excellent reader of the game and an accurate passer, one push of a button will bring up the best options in world football."
between per, kos, gabriel and chambers, arsenal has a good number of defenders. the three main guys (per, kos and gabriel) are interchangeable and play quite well with each other.
now, arsenal 100% needs a defensive midfielder. only coquelin is not enough (arteta and flamini are not good enough).
I wouldn't say that. Our defenders are pretty damned good. Between Per, Kos, Gabriel, Bellerin, and Monreal, I think we're in a pretty good position there.
i found that quote a bit funny. i was hoping that something like this system would be far more useful than that. i mean, football manager (the game) has been used for that purpose for ages now.
also, i don't believe that a system like this one is new. i can't find the source, but sometime ago i read SAP's owner (founder?) had a system much like this one built for the german national team.
The "secret" purchase has been talked about for quite some time now. We'll see if it pays off. In the mean time, while we wait, Manchester United has spent a king's ransom on a teenager and so far it's looking like a great bet (yes, I know it's early still). I'm a strong believer in analytics in basketball and baseball but I wonder if in some sports, like soccer, it is harder or perhaps not possible to quantify a player's effectiveness in a scientific way. I know my skepticism sounds just like what was mentioned in the article.
Having said all of that, Arsene revolutionized scouting when he came to the PL by bringing on somewhat lesser known names from across Europe for prices that proved to be bargains. Some of those players turned into legends. Maybe with a little help he can reinvent scouting again.