2016

Mailbag – Updated Thoughts on Alvaro Morata, Scouting, Manager Evaluation and More

By Ted Knutson | May 19, 2016
2016

Mailbag – Updated Thoughts on Alvaro Morata, Scouting, Manager Evaluation and More

Mailbag time! Readers ask questions. I promise to answer the best of them and not make fun of the worst of them.

I also promise to answer all of these questions honestly to the best of my ability, and not to lie or misdirect people from the truth.

Mostly.

 

build_a_dept

 

Depending on what sort of tools the people you hire bring with them, maybe a month? To some extent it depends on how quickly you get the data company to send you the data, and how much time it takes you to get it organized. As long as I can query a database, we can get work done.

If I walked into a new club tomorrow that worked with Opta data, it would be a matter of days before I could advise on recruitment and maybe a month or two before the Death Star becomes fully operational. On the other hand, any analytics department should spend at least a fraction of their time learning new things so that the value of their output always increases. The curve of starting from scratch is somewhat painful, but as long as you can be less wrong than the teams you are competing against, you have an edge.

Statistical analysis in sport is a competitive endeavour and you will never, ever be smart enough.

Cost is a hella-tricky one right now, and I can tell you this out of personal experience. Part of this is because all of this is new, and football is still confused about the value of the role.

  • A performance analyst role is probably £20-30k on average.
  • Data analysts are not the same as performance analysts.
  • A top data scientist is well into six figures as a base salary in the corporate world. Add some experience to that and you get serious money. (Well, serious for normal people, not for footballers.)
  • Good data people in football should be able to save you millions of pounds a year in the transfer market
  • Smart data people in football should also be able to generate goals on the pitch, provided they are listened to and integrated as part of your process.
  • A single Premier League goal probably equates to £2M next season.

Now consider the fact that in American sports, estimated average salary for General Managers is $2M, depending on sport. (General Manager = Director of Football or "People who can use data to make smarter decisions")

And there’s a pretty strong argument that the good GMs are vastly underpaid. (Read this from Benjamin Morris to explain why.)

Tricky, eh?

If you are willing to hire smart kids right out of school and deal with the mistakes and learning curve alluded to above, you can probably staff a full department for £150k plus data costs. If you want more experience - and there just plain aren’t many people who have experience in stats and have worked in football right now - multiply that by a lot, but you should still get huge value.

 

de_bruyne

 

For those who are unaware, Kevin de Bruyne is one of my favorite players in the world and has been since his first Bundesliga stint at Werder Bremen back in 12-13. When I first started writing about statistical transfers in summer 2013, I suggested him as a cheeky option as a wide attacker for Arsenal, and then thought Chelsea sold him to Wolfsburg for 50% less than they should have. Also two of my three kids are gingers, so “Ginger Freak” is a very affectionate nickname.

Obviously I am still totally in love with the Ginger Freak. He’s one of the best players in the world in transition and this season you can see him learning to become a better player in and around the box. Thinking about what he could become under Pep’s tutelage is the kind of thing that will make City games only air after the watershed next year.

Sorry kids, mature audiences only.

 

young_british_managers

 

Britain has a fairly huge “manager” problem right now. Tactically, the country was backwards for a very long time, especially in the lower leagues. The style of football many of the current 40-60 year old coaches learned is not a great style for the modern game. Thus you are seeing more and more coaches from other countries imported into the Premier League and Championship.

Are they smarter? I don’t know, but for whatever reason, they certainly seem to have had a better football education.

I have a lot of ideas about how to find and educate better coaches – so many that I can’t fit a reasonable answer here. I have also been interacting with smart young coaches a lot in the last year, and I think the next generation of coaches here in the UK will be much more modern and forward-thinking.

If I were a young coach in England right now, I would aggressively try to find internships under coaches that are tactically innovative, because that will give you and edge in the future and the country needs that. You need to learn what they have to teach you, and you need to see how they train those tactics in sessions on the pitch.

If I couldn’t find helpful internships in England, then I would go to Germany or Spain to learn instead.

Coaching is a knowledge-based occupation. If you were going to university, you would seek out the best possible place to teach you. You need to do that with football too, it’s just that instead of a centuries-old university system, you get the chaos and disorder of modern football.

Good luck!

 

younger_player_lower_leagues

 

Rico Henry from Walsall.

English. 18 years old.

League One team of the season.

Left footed.

Plays Left Back.

Zoom zoom.

 

most_impt_thing_at_smartodds

 

For those who have never heard of it, Smartodds is my former employer, and is owned by the same person who owns Brentford and Midtjylland football clubs. The team I was part of sat inside the company, but worked primarily for the clubs – a bit like City Football Group does for Manchester City and their sister clubs in New York, Melbourne, and Japan.

The first project I worked on for Smartodds was probably the most valuable thing I will ever create in football, and I never would have worked on that unless Matthew Benham asked me to do so. It also made me realize how much space exists in football for smart people to innovate.

If you are a team and you want to know what that project was, you can hire me to explain the details. My contact info is at the bottom.

Beyond that, it was fascinating to learn just how hard it is to get transfer deals across the line. Even when you find players early and are the first club to show an interest in a player, and everything looks great the entire time, so many transfers just do not work out.

The other big lesson is that even when you exist in a place surrounded by smart people, plenty of mistakes will be made. It’s part of life, it’s part of science, and it’s part of football.

Learn the lessons and do better next time. It’s so fucking vanilla and self-help it’s almost painful to type, but it’s the truth so there you go.

 

analytics_bigger_topleague_or_bottom

 

In terms of recruitment, analytics creates a bigger multiplier in lower leagues but in real monetary value, just avoiding mistakes in the top leagues is strictly worth more.

To explain: The majority of individual transfers in the top half of the Premier League this season will cost more than what Brentford paid for their entire squad. A team can buy a lot of guys for £500k and sell them on later for £2.5M. You can’t buy a lot of guys for £10M and flip them for £50M though – there just aren’t enough £50M buyers out there (or likely £50M talents).

My team at Smartodds was probably one of the few in the world that did extensive work on transfer evaluations from small and medium leagues throughout the world. Most of the teams that have analysts working on recruitment are in the Champions League and don’t need to look that far down the ladder to find talent. Because we had one of the smallest budgets in the league, we did and we learned a lot because of it.

Number of Championship clubs that have talked to me so far this year about recruitment or anything else: 1.

Seems like plenty of teams are missing some tricks right now.

With regard to everything else stats dorks can do in opposition scouting, style of play, etc, the value is probably equal across leagues, but the competition is worse at the lower levels, so it should be easier to win disproportionately more there on a small budget.

 

what_role_do_scouts_play

 

Tell me what you see on the pitch. What’s the first touch like? How does the player evaluate game situations? What mistakes are being made? What took your breath away, even if it didn’t quite work out? We have a scouting template that highlights the important stuff we want to know, so evaluate that.

Don’t worry about what the stats might have missed. If you explain the football you are seeing in front of you, that stuff should come out anyway.

 

morata_development

 

I thought Morata had an excellent season last year. Adjusting to a new league can be tough, and his performance was still exceptional for a young forward.

On the surface, this season has not gone quite as well, but I’m not sure I agree. Scoring contribution per90 in the league this season was .80. Last year was .88. There no noticeable difference in output except last season he had more goals and this year he had more assists.

He also had 2 goals and 2 assists in 534 Champions League minutes against tough teams, including this piece of filth at Manchester City.

And yet worries (and the narrative) persist. The question you start to ask is why might this season be different than all the others? Stats won’t really tell you the answer to that, so I’d watch every match and compare it to previous years. What’s going on behind the scenes as well? For the numbers being batted around now, teams need to do all their due diligence, almost regardless of whether they think this year was better or worse than Morata’s past.

I thought he was huge value back when he moved to Juventus in 2014. He’ll probably move for silly money this summer and the value will be gone. That said, if I was looking for a guy to lead my line, score goals, and create goals for his teammates as a center forward for the next four years, he’d still probably be it.

For what it’s worth, I still believe Memphis has a reasonable chance to be great as well, but I’m not certain that’s possible under LVG. Football is complicated. So are humans. Memphis only turned 22 in February. Stick him with a head coach that actually wants his forwards to attack and we’ll have a better idea of whether my evaluation was terrible.

 

metrics_to_evaluate_managers

 

There are a myriad of ways to profile head coaches/managers with data and find more objective ways to evaluate them than just the performance on the pitch. It’s actually one of the more valuable services an analytics team can provide to clubs, and there are consultancies out there that quietly do this too.

The big thing to take away from it is whether a particular coach fits the style of play your club wants to play and/or how your personnel might fit into their new style of play. If you have to bin half your squad simply by hiring a new head coach, maybe you want to look a bit closer at some other coaches whose style doesn’t require quite so much immediate, expensive change.

No one writes about it though, because it can be a bit boring and readers seem to care not at fucking all.

 

fav_set_piece

 

*awkward silence*

*clears throat*

Set pieces and their results are random. You might as well ask me about my favorite drop of rain.

*walks out of press conference*

*stops*

Fine, one last question.

 

corner_kicks

 

Eggggzactly. And on that note, I am out of here...

--Ted Knutson @mixedknuts mixedknuts@gmail.com

POSTSCRIPT

post_script