Author Archives: rich

I don’t know what it means, but…

tweaks

This looks at points per game depending on how Kit Symons changed his team.  So in games where we were unchanged under Symons Fulham averaged 0.7 points per game.  In games where we made 1-2 changes we averaged twice as many points.  In games where he made three changes we did better yet, averaging 1.7 points.

I doubt that the differences between the top bars are significant, but there’s something afoot at the bottom.   The same applies when we look at performance v expectation (per previous posts): with no changes the team does worse against opponents than you might hope.  When he made 1-3 changes the team did okay.  (this time, an overhaul caused us to underperform slightly).

Put in language that people might understand, we were unchanged seven times and only won once, 3-2 against a Forest team which pretty much battered us.

In our best games, Bettinelli, Bodurov, Staflylidis, LVC, Parker, Ruiz, McCormack played in all.

I don’t think this has really helped much.

Trying to rank the players

NB – please try to take this in the spirit it’s intended: I’m exploring.  Trying to see where I can get.

You may recall that recently I tried to find a way to objectively assign credit to a football team’s players based on the team’s overall performance. I ran this through the Leeds United team of 1972-73 and was happy enough with the conclusions.

The method is driven by a simple enough ‘engine': in each game, did Fulham perform better against an opponent than would be expected given that opponent’s record across the season? Or worse.

It’s calculated as follows: say Fulham played a team that averaged 10 goals a game and kept a clean sheet. Whoever played in that match would get a +10 for that game defensively, as this attacking giant was kept quiet. If Fulham scored 10 and our team averaged a clean sheet then they’d get +10 there, too, for a total of +20.

If Fulham had lost 10-0 then this would have been a zero, as Fulham kept the opponent exactly to their seasonal average.

We add up the totals for the season and see which players played when opposing teams were kept above or below seasonal tallies.

Phil Magnus took my Leeds spreadsheet and filled in the latest Fulham season. If we apply the above literally we get the following ranking of players:

Christensen
Ruiz
Burn
Bettinelli
Rodallega
Stafylidis
Parker
Hoogland
McCormack
Hutchinson
Bodurov
Tunnifcliffe

This is all players who played 20 games or more. Which is to say that the team performed MUCH better with Christensen in the side than without him, and a fair bit worse with Tunnicliffe in the side than without him.

Okay, the elephant in the room is McCormack, who played so much he had to take one for the team, which is to say that we couldn’t get a with or without you for McCormack because there’s no without you part. He’s largely alone in this, although Betinelli (39 games) does well because Kiraly and Joronen had bad happenings when he was absent.

Among non-qualifiers Elsad Zverotic (5 games) was actually top, having played his part during a relatively unspectacular but almost unbeaten run halfway through the season. Turner was up there between Ruiz and Burn, and we did okay in games when Richards played. The team did badly in games when Fofana started and also Woodrow, Amorebieta and Hydman (Magath effect!).

I did a second version of this in which defenders got more credit for defensive performance and attackers more credit for attacking performance. So, if we lost a game 6-4 the defenders who played would get fried and the forwards would get credit. That’s probably reasonable enough.

For context, I put these numbers up against Whoscored.com’s ratings and against a subjective view from Mike Gregg, who is good at these things and, I find, takes a balanced approach that neither swings too far towards the ‘populist view’ nor tries to be too contrarian (which I, of course, am prone to do).  These are all rankings. The numbers are meaningless in this context, really:

umbers

RAW is just the pure data: how did the team do when he was playing?  POS is when I adjust this for position played; WS is WhoScored; MJG is Mike.

Hmmm. I think Whoscored is too high on our defenders (top ratings generally going to these people) which can’t be right under the circumstances. Mike’s views, I think, cover a consensus quite well.

So I don’t know. My initial ratings are essentially facts (facts that are open to all kinds of distortion and factors beyond folks’ control, but facts regardless.. kind of): when LVC played, teams did worse against Fulham than they did against all other opponents. Ruiz and Burn had similar impacts (as did Turner later on). You can slice this around a bit to try to account for responsibility and maybe that works, too. Not sure. Clearly I have a way to go here. But the key messages are:

Christensen is really good
Dan Burn did about as well as anyone could in that defence (which is a slightly circular argument, I realise)
But both look like the pillars on which this team needs to be rebuilt
If you take out the Whoscored defensive bias (?) then Hoogland, Hutchinson and Bodurov get panned. Ryan Tunnicliffe doesn’t come out of any of this well.
McCormack was unfairly held back by the raw approach to all this because he played so much, so he must be seen as a success (duh!)
Rodallega’s probably a fair bit better than the fans give him credit for
Parker did okayish

How Matt Smith explains the universe (more on how we can bounce back)

One of the central thrusts of recent posts has been how little of a difference single players make to teams. I’ve currently settled on a range of about -5 to +5 goals per season per player, which doesn’t sound like much but which, really, is about all that makes sense (see previous posts for more).

A nifty example of all this lies with Matt Smith, widely acknowledged as having been a great success with Bristol City this year.

Smith scored 13 in 17 games and his manager at the time was extremely positive about his contribution to the team.

Thing is, while Smith was at Bristol the team averaged 2 goals per game to .8 conceded, and while Smith wasn’t at Bristol the team averaged 2.1 goals per game and .8 conceded.

Pro-rata that over a whole season and you get a +59 goal difference without him and a +54 with him. Both are phenomenal, and I’m not about to suggest that the team were actually better off without him, but the point stands: even at this level of productivity, the team didn’t miss a beat without Smith. They worked as a team, and while he was there Smith was good enough to turn this overall play into tangible results; when he wasn’t someone else was.

mattsmith

(There is some substance to what we might call a ‘launch pad’ theory here: Bristol were good but not superhuman before Smith, got really good with Smith, then were astonishing after he’d left. This is without correcting for opposition, and is skewed by two thrashing late in the season, but those thrashings did happen: you don’t win 8-2 or 6-0 without being a really good team. So.)

This, I think, brings us back to the coaching. A good coach builds a system to the extent that it doesn’t necessarily matter which players are on board on any given day. We saw this with Roy Hodgson, and while few managers build system-based squads to this extent, you need to be some way in this direction if you’re ever going to succeed: managers who insist on keeping on changing personnel, looking for “the right combination” are sometimes, I feel, just deluding themselves. The right combination is always just around the corner, if only the right player can be secured. This is true to a degree, but again we come back to to point we’ve made a few times this week: you need a solid tactical base (you need to put players in a position to succeed!) and solid recruitment.

Fulham need a platform and a profile. It’s become too scattergun, too random. We need to get back an ethos, and build a team around this. This, I feel, is crucial to our bounce-back, if it’s to happen.

Alright, maybe it is possible: how Fulham can bounce back

wall

Thank you to those of you who took yesterday’s post in the spirit it was intended.  The idea is not to literally claim that we need players of a certain absolute quality, more to point out that there’s a long way to go between here and there.  The comments section was suitably thought provoking, which leads me nicely onto why we may be better able to bounce back than I think.

First, this notion of a +5/-5 player relates to what that player accomplishes on the pitch. So a -2 player might be a +2 player in other contexts.  That other context here cuts two ways:

one, if we have a new manager, or if Kit gets a full head of steam, can this player contribute to a good Fulham team?  Put another way, is the player better than he’s shown this season?

two, if the answer to the above is no, is there reason to believe that the player can mature into the player we need him to be?

And here’s where things get more encouraging.

Having watched a lot less of Fulham than most of you this season, here are a few thoughts to start things off:

Good enough to make Fulham effective again: Christensen, McCormack.  I think these two speak for themselves. McCormack lacks for nothing at this level, Christensen emerged as a potential star.

Good enough to be part of Fulham’s next effective team: Bettinelli, Burn, Kacaniklic, Smith, Rodallega. I might be overreacting to Bettinelli, who seems to have had one of those superficially spectacular seasons whilst conceding bucketloads of goals.  Hunch is that he’s done about as well as anyone could have been expected to this season and that the issues are largely in front of him. Burn gets a pass having demostrated some ability at this level before and indeed, to a lesser degree, in the Premiership.  He’s not the finished article but I feel the materials are there.  Kacaniklic – and I appreciate I’m in a minority here – is a Premiership player so ought to be more successful than has been the case here.  He’s been messed about no end and needs a defined role and some patience. Smith I like, and have been gratified by his performance this year.  No, he might not be your first choice striker in a promotion chasing team, but he’s good enough to be in the conversation.  As is Rodallega, who can happily function at a higher level and therefore presumably just needs better teammates/managers/attitude?

Not in either of the above groups but shows potential/some evidence that might be improvable: Joronen, Hutchinson, Bodurov, Kavanagh, Grimmer, Tunnicliffe, Hyndman, Williams R, Roberts, Woodrow, Taggart, Dembele, Burgess, Eisfeld, David, Mitroglou, Williams G.  More on this group in a minute.  Some real talent in here, which is encouraging for our purposes.

Not in any of these groups: Kiraly, Hoogland, Stafylidis, Voser, Parker, Amorebieta.  I’m being harsh here because Fulham aren’t going to progress by not being harsh.

The missing link in all this is the manager.  We’ve talked about needing a 40 goal swing next season and as people pointed out in the comments below, that absolutely can happen.  A bit of positive momentum and the squad we’ve just itemised above can look rejuvenated in no time.  So while we have castigated the team all season, there’s some talent in the group.  It just needs unlocking.

Can a new manager do this?  Absolutely.  The current manager *might* be able to.  Momentum’s everything: with a good start and some new ideas, anything might be possible.  The season’s a bit too long to fluke anything (unless you’re Derby, for whom the fluke worked backwards) but for sure things can get better.

What does it take then?

First, we need to address the whole “diabolical defence” thing. I think that means a commitment to 4-4-2 until we’re good enough to grow out of it.  No, it’s not sexy in this day and age, but for as long as I can remember we’ve wondered whether Scott Parker can protect a back four (no); whatever happened to the midfield protecting the back four as a group? Two banks of four, hard to beat, then let the McCormack and Smith combination do its bit as and when it can.

I do think this means we need to buy the best centre-half we can lay our hands on. We need two very good full-backs whose first job is to defend, and from there we might be able to muddle through with a midfield built around LVC, Tunnicliffe, Kacaniklic and an up and comer on the right learning his trade. If Kacaniklic doesn’t work back enough he can’t play; from now on it’s about teamwork.  The team is incentivised on clean-sheets, the team becomes about clean-sheets. Nobody wants to play Fulham. That kind of thing.  This Fulham side is far, far too easy to play against: those Norwich goals were just the latest in a constant stream of dreadfully defended concessions, whereby opponents entering our area were neither closed down, tracked, or any of those things that might make it harder to score.  You don’t concede nearly 90 times in a season with any idea what you’re doing back there.

So yes, maybe I can see grounds for optimism, and maybe it doesn’t need millions of pounds either. But it’s about time Fulham stopped just throwing muck at the wall and hoping some of it sticks. It’s time to address the very obvious weaknesses, not by constantly trying new players and hoping they’re the answer, but by making a real commitment to doing things better.

There’s no coming back from this: why I don’t think Fulham can achieve promotion any time soon

ks

Based on the final analysis it looks like Fulham will end with a goal difference of about -20, which means they’re something like 5 net goals above the relegation places.

Looking at the table, the promoted clubs had goal differences of 41 and 50, which we’ll call 45, and the playoff clubs are 19, 31, 32 and 38, which we’ll call 30.

In short, next year Fulham need to be a +30 team for the playoffs and a +45 team for promotion, from a base of -19 (this season).

lt

That’s going to be a +50 odd swing for playoffs and a +65 swing for promotion!

The other day we figured that a really good player might be worth 5 net goals to his team.

You might think of footballers on a 10 point scale, ranging from -5 to +5. At one end you have a team that, if comprised of players of this quality, would end with a goal difference of -50, e.g. Blackpool. At the other end you’d have a team romping away with the league at +50, e.g. Bournemouth.

If we think Ross McCormack really is a top quality player, a +5, then his teammates are most likely a bunch of -1 to -3 players. Nothing else adds up. I mean, we can argue that the players are better than they performed, which might be reasonable, but ultimately good players play well. Fulham’s players didn’t play well, so it’s hard to say they’re good players.

If I’m going to keep this mathematical, Fulham need to turn the -1 and -2 players into +1 and +2 players. To make the requisite jump you’re looking for a 50 goal swing remember, either not conceded or scored.

You’re either having to improve players who almost got relegated into a promotion machine (how?), or you have to find the equivalent of 10 more players who can all play at a level that’s 4/5 net goals better than the player they replaced. I don’t think that’s really possible.

In any case there are two possibilities:

1) better coaching can radically transform these players
2) we need a complete overhaul and even then need to spend massively and effectively

Does point 1 work? As best I can tell there aren’t many managers who have made a career out of turning water into wine. Most managers achieve what they should with what they have. Now, we come back to this argument that the squad significantly under-achieved, and I’m not completely against this idea, but it seems more likely to me that the players simply weren’t very good.  Our eyes might see promise there but the goals against totals suggest that something’s up.  That the team was unable to stem the flow at all suggests that either the coaching was awful or the players didn’t have it in them, probably both.  Either way, if you want a +50 swing in goal difference it’s not going to happen by magic.

Does point 2 work? No. To acquire that many good players, without misjudging any of these acquisitions, is too big an ask.  You can’t just buy 9 more Ross McCormacks.  You couldn’t afford to even if you could find them all.

As things stand I simply can’t see how this squad can become the squad it’s meant to be. To believe otherwise seems to me too great concession to wishful thinking.  It’s too much.

How important can a single player be?

One of the things I always wonder about in football is how important any given player really is. When you think about it, there are only so many points to go around.

First, a disclaimer: this is all theoretical. I know, I really do know, that football is a team game. That the interactions between players are complicated, that assigning individual value to a player transcends this kind of analysis. That managers sign players to fit into a role in a system, and that this system is often what makes a team win or lose. I get all that. This said, players must have some inherent value, right? Or why buy them? If you pay whatever million for Ross McCormack it’s with an expectation that he will make the team better. My question, then, is how much difference can one player make?

For reasons we’ll get to in a minute, take a look at the 1972-73 season. That year Liverpool won the league with what would have been an 85 point haul (2 pts for a win then). WBA were bottom with what would have been a 37 point haul. So over 42 games the best team in the league were just over a point a game better than the worst team. Put another way, Liverpool scored 34 more goals than WBA and allowed 20 less, so you’ve effectively got a 54 goal swing between the best team in the league and the worst.

table

So over 42 games the best 11 players in football are combining to be 54 goals better than the worst 11 players in football! If you want to divide this up and assume 11 players played all the time, we might assert that the very best players are worth 5 goals a season more than the very worst players!

Think about that for a moment.

In baseball they call “the worst players” replacement level. Simply put, it is assumed that most first division teams could field a team as good as the worst team in the division, so what you’re looking at is a concept of value over a replacement player. Here we see that to be the best team in the league you need to find 55 net goals over an awful team. So, on average, your players need to be worth five goals a season above bad. That doesn’t seem like much, but really, nothing else adds up. Sure, it may be that this 55 net goals breaks down as 15 goals above replacement for a superstar and everyone else four goals above replacement, but you get the general drift. The point is: there’s only so much value to go around. Unless something very odd happens, most players can only ever have incremental value.

For reasons best known to me I dug deep into that 72-73 season, focusing on Leeds United. I looked at all the games, looked at how Leeds did, how good their opponents were, and which Leeds players were playing at the time.

What I did then was to assign credit for good performances and debit for bad ones. So against Liverpool that season, if Leeds held Liverpool to fewer goals than Liverpool usually scored they got defensive credit, if Leeds scored more than Liverpool would usually allow, Leeds got attacking credit.

Leeds also finished 50 goals above WBA, so let’s divide up those 50 goals based on the digging I’ve hinted at above (I won’t go into the details). We get this:

Leeds

What? All it is really is apportioning the surplus points based on how the team did when each player was on the pitch. So here we have a number of players who were on the pitch when Leeds did well. John Giles by this method was the most important player per appearance, Leeds doing best of all when he played, but he only played 33 games. The numbers respectively are those 33 appearances, 27.7 means that when Giles played opponents were held 27.7 goals (for/against remember) below what they achieved in all other matches that season, and pro-rated is just fitting all of this into our 50 goal surplus (how much better Leeds were than the worst team in the league).  When Giles didn’t play Leeds underperformed a bit, so players like Peter Lorimer, who played most games, don’t get much more overall credit than Giles. If Giles had played 42 games he’d be the clear leader.

Anyway, by dividing things up we get a sense of who contributed what. True, we can’t untangle exactly who did what, but we have some clues: when Roy Ellam played, Leeds struggled mightily. This wasn’t likely all Roy’s fault, but he played in what were otherwise full-strength teams. They lost 4-0 to Chelsea, 2-0 to Liverpool and 2-1 to Birmingham in those three games, as well as drawing 2-2 with Palace. Given that Leeds were arguably the division’s best team that year, we can see that Roy’s presence was not conducive to Championship form. It might have had nothing to do with Roy but I have no way to untangle that.

Otherwise most of Leeds regulars might be said to have contributed about 5 goals above that replacement level. I appreciate that this is fudged slightly because I set 50 as the total ‘pie’ to share out, but hopefully by working through a single team in detail we can see how – almost by definition – no single player can have *that* much impact on a season. There are only so many points to go around, even on great teams. If you think Allan Clarke’s worth 20 points a season to the team, if you think Giles and Lorimer and other superstars are similarly gifted, then before long you realise that you’ve run out of points, and that the likes of Billy Bremner and Jack Charlton and Norman Hunter, well they’re worth a piece of the pie, too, right? And so on and so on… So you come back to where you started, that even the best players in this team may only have contributed 5 net (attacking or defending) goals more than an average player might have done in their place. Nothing else adds up.

Right, if by some miracle you’re still reading, you’re wondering if this is the right way to look at things. It probably isn’t. But again, when we buy a footballer we assign a value to him, an expectation. And I’m saying that these expectations almost have to be too high. There’s only so much surplus value you can fit into a football team.

This affects the way all teams do business. If Fulham needed 50 net goals to be competitive this year (and actually the Championship still seems to broadly support this) then they had to understand a) what they had and b) what they needed. Ross McCormack brought all those goals with him, and we had a good wonder about what that might mean for Fulham, but in retrospect we needed more players of that quality, because a single star player on his own isn’t going to cut it.

The Secret Footballer on the Championship, and how Sam Allardyce’s analytics made a difference

davies460

More brilliant stuff from the Secret Footballer about how Sam Allardyce used analytics to his advantage:

Allardyce’s analytical ideas were impressive. He looked at everything, every component of the game.

At one point, he stationed a player at attacking corners in an area of the pitch that appeared to be in the middle of nowhere.

His logic was based on seasons of statistical research that determined where the clearing header was most likely to land on average.

The clearing header is the most likely result from a corner so it pays to know where it’s most likely to land.

The reason for that is because the stats for goals from a corner go down with every recycled ball.

It goes something like this. The corner is aimed at the middle of the goal.

Statistically, this is the area, actually just slightly towards the near post, where most direct goals, usually headers, are scored.

But there are stats at work here, within stats.

Because if that header isn’t a shot, then the next area that it is likely to end up in is the back post or “Pomo” – the position of maximum opportunity.

Allardyce would station a man here, running in on the outside of the back post for what would be a tap in. Very often this was Kevin Davies.

The third part of the area that the ball would end up would be just in front of the goalkeeper and you can still see Allardyce’s stats at work here. This time, with West Ham United.

In front of the keeper is exactly where Kevin Nolan stands and I’ve lost count of how many goals he’s scored from this position.

All of these three parts of the penalty area are dependent on a Bolton player winning the first header. After that, the stats go down markedly.

After that, the stats in play are for a recycled ball … or the second phase of play.

And given that the clearing header is generally towards the dugout near the halfway line, there are two options when keeping the attack going.

Either hit the ball to the back post, where players like Davies will knock the ball down for someone to shoot, or pass it to the corner taker, who hits an inswinging cross, usually nodded home by a centre half or a striker.

People thought that Bolton were a big, cumbersome team. They weren’t; they just played the stats better than anybody else.

He’s likely to talk about Fulham tomorrow, which should be interesting to say the least. It’s a fascinating series so far.

Scouting for Fulham part 3 – tying it all up

Some of you will have noted my starting with League One in this series. Why?  Well, simple: the method is looking for the very best players by looking at elements of teams that are performing exceptionally well relative to other teams.  Those teams will usually be towards the top of the league, so is it realistic, in our position, to be cherry picking the best players from the best teams? Not at all. But you can probably do that with the best players in League One.

Anyway, all well and good.  But where does this leave us?  A load of names that might or might not lead to anything.  Thing is, we can easily back check the method. It’s a manual process so I’m not about to do to much, but in the interests of validation I’ll do the Championship seasons 2010-11 and 2011-12.

Here’s what the method would have led me too.  In short, it’s effectively an approach that says “Dear Mr Scout: watch x team and pay attention to players in y positions”.  So:

2010-11
Norwich: Grant Holt, Wes Hoolahan
QPR: Kyle Walker
Swansea: Ashley Williams, Neil Taylor, Angel Rangel, Joe Allen

I’m really happy with this group. Holt might not be everyone’s cup of tea but he was certainly effective. Hoolahan’s a terrific player. Any of those Swansea players would have been a big asset to Fulham.

2011-12
Forest: Wes Morgan (now Leicester captain)
Southampton: Rickie Lambert, Adam Lallana
West Ham: Kevin Nolan, Ricardo Vaz Te, Carlton Cole

Same here.  The West Ham cohort will raise the odd eyebrow but these are effective players, and the point here is that the system was looking for the best players in the Championship – it’s hard to argue that these players were not among the best players in the Championship.  The Saints players were clearly legit, and Wes Morgan probably would have represented something of a diamond in the rough.

I’m cherry picking after the fact but not by much. The whole point here is, as noted above, to provide scouts with areas to explore thoroughly. Any scout told to go and watch Southampton and to keep and eye on the forwards would have come back with glowing reports on the above players. Ditto Swansea’s defenders.  It works.

So there’s no reason you couldn’t do this for Sweden, for instance, or Denmark. If I may name drop for a second, while writing my Roy book (four paperbacks left – or find it on Amazon!), Erik Nevland really stressed to me the importance Hodgson and Lewington placed on professionalism, on character.  It’s no coincidence that they dipped into Scandinavia so often.  There’s no reason we couldn’t do this now.  We have a fair idea that the method works in identifying the better players, but we also know that we’d be paying a premium for the kinds of names we’ve been finding.  But dig around in Scandinavia, focus your search on the parts of teams that are excelling in their particular job, scout them intensively (note: you cannot, cannot, cannot, do this kind of thing with analytics only; you need every kind of information available to you, and attempts to create false dichotomies are a waste of brain-space) and see where it leads you.

This isn’t particularly sophisticated analysis, but it’s searching with a point.  Also during my research for the Roy book I found some really good quotes from the WBA chairman, who, after hiring Hodgson and director of football Dan Ashworth, was scathing about how English clubs do business:

“It reached a watershed seven years ago under Bryan Robson when we were bringing in older players who had maybe had their day. I thought to myself: ‘If I’m spending the club’s resource, based upon recommendations made by people who might not be here in a couple of years’ time – for whatever reason – and we are left with those problems, I’d rather make the decision driven by the right methodology rather than on a whim. Basically, if I’m going to make a mistake, I’d rather make it myself.’”

So he went around Europe to see how other clubs did it. They found a role – sporting director – which English clubs really didn’t like, but which European sides felt was all but essential. “They were all looking at England saying: ‘We cannot understand what is going on when we deal with England. The clubs there pay top price, they don’t really check what they are buying.’”

Hmm. If that doesn’t resonate, it certainly should.  This is why there’s a need to think about rigour, about processes, and yes, about backing up some of these decisions with analysis, not just the whims of people who might or might not be good judges.

Look, analytics is a dirty word among football supporters.  But it was among supporters in other sports, and in those sports, the teams that have failed to embrace all available sources of information have fallen behind.  Literally every misgiving you see about statistics on fans’ sites is not a reason not to do this.  Yes, football is complicated, yes you can prove anything with stats, yes you can mis-read stats, but that’s why you have some of the brightest people in the world moving into these fields, to guard against exactly this (against this, Sir Alex Ferguson apparently sold Jaap Stam because he misinterpreted some tackling stats – it happens!).  It’s a slow process, a hard one, and even when clear truisms are being found, getting buy in from the football side of things won’t be easy.  But any forward thinking club needs to do this.  It doesn’t cost much, but the rewards can be great.

Scouting for Fulham part 2

Okay. Well yesterday we used the power of high level analysis to try to find some players. I had to use some quite high level numbers because that’s all I have, but the thinking felt sound: identify teams who do a certain thing extremely well then try to identify which player might be behind this. Did it work? Who knows, but I enjoyed the process, so I’m back again today with the Championship.

I use standard deviations, which is a statistical measure of dispersion: the further away from the average you get the more unusual the skill. Generally speaking, anything over 2 SDs is pretty amazing. So:

champ

Bournemouth are two standard deviations above average in scoring, shooting and getting shots on target. It’s clearly an attacking setup so hard to pull out individuals, but to be fair the defence is doing pretty well too. Looks like a fine all around team, well coached, and unmistakably for real. How do you pick a player or players blind like this? You can’t.

Middlesbrough are two SDs above average in not conceding, and in not allowing shots on target. They are not quite so good in preventing shots overall (still very good) which suggests that teams rarely get a good sight of goal and that there are men behind the ball doing a very effective job. So, as yesterday, we’ll look for players 27 or under who are in the right areas.

George Friend 27D, recently player of the year (12-13), signed from Doncaster (where he was supporters’ player of the year) with Ipswich and Forest wanting him.
Adam Clayton 26M, Manchester City from age of 7, Leeds in the pedigree, England U20, signed from Huddersfield last year, where he had also been that club’s player of the year (someone found 12 minutes worth of highlights!
Ben Gibson 22D North East Writers Player of the Year 2014, up through the ranks
Daniel Ayala 24D much touted as a youngster at Liverpool, now apparently playing extremely well for Boro

Norwich look awesome by these measures, good at scoring and shooting, and hardly allowing anyone to shoot against them (2.5 SD above average). The fact that they are shooting a good amount suggests a near absolute control of football matches.

Bradley Johnson 27DM
Nathan Redmond 21AM
Martin Olsson 26D/M
Jonny Howson 26MC

All of the above have been around the block a bit, playing at the top level and with some plaudits. Redmond well hyped, of course.

Watford don’t shoot that often but do score often. Top finishing or expert chance creation? Either way, some top notch front play.

Troy Deeney 26F has scored at a goal every other game for three years.
Odion Ighalo 25F I’m in love with Ighalo. He looks like a faster Clint Dempsey, apparently capable of scoring any type of goal you might imagine. . From the Udinese talent pipeline.
Matej Vydra 22AM aslo via Udinese, Czech international, another lovely player.

And we can’t do this without mentioning Fulham.  If we were scouting Fulham we’d note that the attacking element is passable but defensively less so, to the point where nobody at all in the division has allowed more shots on target.  That’s terribly damning: not awful in preventing shots overall, but give away really good chances, suggesting attackers not being picked up or closed down.  There’s nothing here that Fulham do well, so as an analyst, searching for hidden value (alright, hardly hidden), you’d take a quick look at those Fulham numbers and move onto the next team without dwelling for long.

Scouting for Fulham

Tony Khan, the owner’s son and statistical head honcho for the Jacksonville Jaguars, crashed into my world this evening via Twitter.

tonykhan

Well.  Given our current status we’re probably going to have to find players from leagues below our own, so that’s hard, because there’s not much data on League One.  Never mind, I thought, let’s see what we can do.

You will by now be aware of the general perception that football is about goals for and against (I’ve prattled on about goal difference being a good indicator for years, yes?) but also worth noting that shots, and particularly shots on target, tend to be even better.  And obviously there are two sides to this coin, those shots you take and those you allow.  So:

SDscores

Hehe. So, what we’re looking at here is how the teams in League One do at scoring, not conceding, shooting, not allowing shots, getting good shots, not allowing good shots.  We then compare the teams to the rest of the league and highlight situations where a team is an outlier.  As an example of this, teams only get one shot on target for every three shots they take against Preston, which is sensational and suggests that it’s almost impossible to get off a clean shot against them.  This could be because of their defence or their midfield, but in any case, it’s what we’ll call a lead.  Below we explore some leads.

Bristol C better (outstanding) F/A than indicators suggest = v strong midfield?  Hard to know though, seem to be excellent everywhere so not clear who to spotlight.

Crewe half of their shots are on target, much better than rest of division.  Suggests creating v good opportunities and/or good forward play. So:

Nicky Ajose 23F Didn’t make it at United, but Ferguson sold to his son Darren at Peterborough suggesting Fergie saw something. On emergency loan from Leeds.
Anthony Grant 27M Chelsea youngster, made first team squad in 2005-06, no appearances.  England U16,U17,U19. Last year made available for loan by manager owing to attitude issues.
Burton+Albion+v+Stoke+City+gDahRnr9aITl

Jamie Ness (above) 24M – young Rangers prospect who had injury issues and who left when the club went into administration.  Signed for Stoke but didn’t break through. Been loaned out to Orient and now Crewe.

What I’m doing here is scouring the team’s stats for players who play regularly in the positions we’re interested in and who are under 27 years of age.  They might be rubbish but you have to start all this somewhere, right?  I’m delighted with my first three stabs, all three players having significant pedigree in the game.

Fleetwood – their opponents need a lot of chances to get one on target. Fleetwood concede many more shots than they take, but similar end up with similar numbers on target. e.g. they are much more efficient.  Suggests perhaps a good counter-attacking team?  Not sure. In any case, they have two young full-backs who seem interesting:
Josh+Morris+Huddersfield+Town+v+Blackburn+hhegno8xEfgl

Josh Morris (above) 23M young left back on loan from Blackburn, where he was well thought of.
Conor McLaughlin 23D young NI international
Antoni Sarcevic 23M Man City youth from 7 to 15, fell away, played non-league, now working his way back up.

Preston have outstanding defence. Almost 3 Standard Deviationss above average for how hard it is to get a shot on target, which is nuts.

Paul Huntington (below) 27D centre-back, former Newcastle player.  I’m going out on a limb here: a centre-back, aged 27, came through with Newcastle, now anchoring one of the toughest defences in League One: I reckon Huntington’s got something going for him.

Paul+Huntington+Preston+North+End+v+Blackpool+0nl-n3eD17Bl
Tom Clarke 27D defender or midfielder, former England youth
Joe Garner (below) 26F – the thing here is that Preston have these insane defensive stats but Garner’s only gone and scored 20 in 29 games. Prolific all his career but hasn’t really stuck. Often a victim of numbers, e.g. signed for Forest for £1m but stuck behind a number of decent options. Worth a punt.

joeg
Bailey Wright (below) 22D Preston young player of the year in 2013. Australian youngster.

Bailey+Wright+Australian+Socceroos+Headshots+P9IMhoSG_WVl

Conclusion

Look, I’m not an idiot.  I know that you can’t just pluck names out of thin air, copy a paragraph from wikipedia and proclaim yourself the Billy Beane of football, but there’s method to the madness.

1) we’re finding teams who are outstanding at a particular facet of their game. This is important: it gives us some assurance that we’re finding players who can do a job.  We’re not being blinded by perceptions, we’re finding defenders who are part of teams who are great at defending, for instance.  We can’t know from here why the teams are so good at defending, but by looking at the defenders who have played most often, not been a sub, etc, we can have a fair guess.  It is only a guess, but an educated one.
2) we seem to have stumbled upon a number of players with a decent pedigree.  This is suggestive of talent, which is a big deal.  It’s very easy for footballers to get lost in senior football, there’s so much luck involved in who makes it and who doesn’t (I really believe that).  So there’s little doubt in my mind that there are gems to be found down the leagues (or how do you explain Bournemouth or Brentford?).  I think this is one way to find them.3) clearly the next step would be to get some qualitative feedback.  That’s exactly what I’ll try to do next.

But I wouldn’t be that surprised if taking this kind of iterative approach to building shortlists might not be the way forward.  As I’ve discussed too many times to mention, our eyes are notoriously poor judges of anything, and if players could be effectively judged this way all managers would agree who the best players are and nobody would make mistakes in the transfer market.  This doesn’t happen.  When you break things down to actual achievement then you’re not letting prejudices blind you; you’re finding defenders who have had success defending.  Which is what you want.  Then of course you have to untangle the information you get, but it’s a start, right?