By Tony Attwood
On 2 September we bade farewell to the great Arsenal Transfer Index – a listing of all the players we could find who were allegedly leaving the club or coming to the club. In our final analysis we noted
- 115 incoming transfer rumours – almost all of them wrong
- 22 players who were about to leave Arsenal – most are still with us.
- A total failure to spot the actual transfers ahead of activity – except in one case.
and I had a scribbled note on the back of an envelope I have now lost, to say that perhaps 1 November would be the date to start again.
But really there just aren’t that many rumours around, and most of those that can be found are so plain daft that you’d think I made them up.
However there are one or two bits of gibberish floating in the ether – although they are all very different from the summer. Primarily the talk is of full backs. We need someone new to cover for Bellerin – preferably an international who will keep his excellent talents well honed even though he is not playing every week, or at all. And someone at left back because Monreal has utterly lost it and Gibbs was rubbish against the Bulgars.
What’s interesting in such chitchat is that none of the issues that arose last season appear to have be learned. Issues such as
1: You can’t get a top, top player as a back up and expect him to live a life on the beach.
2: Just because a player has one or two duff games, that does not mean he is useless and should be jettisoned. Jenkinson hasn’t played for a long old time, and when he was playing before it was for a bunch of mobsters who are more attuned to publishing pornography, denying reality and taking tax payers money than anything actually to do with football.
3: You can’t judge on one game. Gibbs has only played bit parts and there is hardly a player in the country who can move from that position into top class performance. Buy in a new player to replace Gibbs and you then tell Gibbs it is time to move on. So that leaves the problem of finding another backup player and we are back into the whole cycle over and over again. Besides not all transfers work (although in the mythology of these people if only we got rid of Wenger all transfers would then work, because he is the ONLY manager who ever buys a player who doesn’t make it.)
4: In short buying a new right and left back as some sites are suggesting today is fraught with danger and difficulty, and we would surely be much better off getting Gibbs and Jenkinson, and come to that Debuchy, back to full strength and ability. While at the same time see who else we have got in the under 23s who could fill either slot as a backup.
These issues led me to wonder, who do we actually need now? In the summer we had the orchestrated “Spend some” campaign which didn’t change anyone’s mind about anything and the desperate hyperbole about buying a centre forward and central defenders resulted in buying two players who a) no one had tipped to be coming to Arsenal and b) not too many people had een heard of.
What we also discovered with all the shouting was the Giroud was useless and would never score goals. Except when he does. What we didn’t learn was that Alexis was doing as well as Henry in terms of two year worth of goalscoring. What we learned from the press (by investigating their wholly misleading articles) was that Arsenal were one of only five teams that had two players who scored 10 plus in the season. What we have now learned is that we are the only team with two players already edging towards double figures.
Which brings me more or less to my point. The summer was full of “Arsenal need a central defender and a centre forward”. So what will the winter be full of? Two full backs? Or… will we have new players coming through, not in terms of transfers, but in terms of stepping up from the youth team.
What, for example, about Maitland-Niles as our next man to step up? In September the bizarre Pain in the Arsenal website proclaimed that “Ainsley Maitland-Niles Arsenal career already appears to be in turmoil after no loan move materialised for him in the summer transfer window.
“Ainsley Maitland-Niles is a player who was once, and still is being tipped to be a future Arsenal great. However, despite a promising loan spell at Ipswich Town last season, failure to grab another loan move this time out might see his Arsenal career falter and then eventually die.”
Fortunately the Metro was a little more in touch with reality with “Arsenal ace Ainsley Maitland-Niles scores brilliant goal for England’s U20s vs Brazil” as they realised that the management saw Maitland-Niles’ best chance of greatness as being under the direct tutelage of a man who knows a thing of six about bringing players through. N-M started the game against Forest at right back this season – before also starting the Reading game in midfield.
Bringing players through in this way is the ideal – Jeff Reine Adelaide seems to be highly tipped at the moment although as the squad gets stronger it gets ever harder for players to push up from the under 23 team.
Certainly we seem to have a multiplicity of goalkeepers, midfielders, wingers and centre forwards at the moment, leaving me to wonder if we will have another transfer window denounced by the hooligans of the blogsphere because “Arsenal was the only club in the galaxy not to spend £100m+ this January.”
Of course the clubs that have not filled up their list of 25 will be trawling around looking for “home grown” players to purchase at any cost. Maybe our biggest task will be to hold on to the youngsters we’ve already got.
January is often our month of talent scouting for the academy, so I wouldn’t be surprised if we buy 1-2 under 18 players from Europe. And given that our U-18 and U-23 are bit short on the defensive side, we might see someone coming in.
As for the current young guns mentioned above, AW is slowly integrating them in the squad, in hope of them becoming part of like Bellerin or Iwobi or Coquelin.
OT:
Sorry. I now have Wenger’s 20 years (plus the first few games of 1996 before that) sitting in a program. Any (other) questions you want answered?
There are 34 sets of contiguous 4 day sum of points for a season, times 20 seasons is 680. Plus a few games in this season, so almost 700 data points.
Yes, we could need to buy a senior player for our senior squad team if the need for it arose.
So far, our season campaign is moving fine as we are still in all titles contest contesting for honours.
God forbids, an injury occurring to a very important Gunner in our senior team squad that sidelines him for months, could force Le Prof’s hand to seek for external solution if the problem defiled internal solution. So that Arsenal title charge in all fronts will not suffer any setback but keep moving on to the right direction as planned.
However, the indication is, with the prevailing injuries to some of our current regular playing and backup Gunners such as, Cazorla, Walcott, Bellerin & Monreal(all short term absentees) and a backup player – Lucas(recently sideline for 8 weeks), Arsenal have won a PL game and a Ucl game without the services of these top Gunners.
Arsenal have cope well without dropping points even in the face of this arising mini injury crisis in the senior team squad which suggests Arsenal strength lie in depth.
The stuffing of our senior team squad externally and internally with top quality players this season by Le Prof has made Arsenal to be sufficient in strength and backups. Therefore, a need to enter the market may not afterall becomes necessary for Le Prof to undertake in January.
OT.
With this night’s 0-1 home lose to Bayen Leverkusen by Tottenham HS in the Ucl match at Wembley, is the stage set for a crulesome match between Arsenal and Spurs on Sunday at the Ems?
Will Tottenham HS want to use this match to bounce back to winnings way now that their unbeaten run in all competitions has gone up in smoke?
I doubt seriously if Tottenham HS can bounce back to winnings ways at the Ems on Sunday after the recent setback suffered by them to their campaign in the title race.
Nevertheless, I want Le Prof and the Gunners to take heed of Spurs coming to the Ems to spoil the party for us by fight like a Lion that is wounded and thus put adequate the necessary measure in place to deny Spurs coming out from the Emirates Stadium with a point which they wouldn’t mind to get.
Controversial post coming up…
Did we need to buy anyone in the summer?
From what I have seen so far, I don’t believe any of our summer purchases have improved the first team. A greater strength in depth, yes. Injury cover, yes. But a clear improvement? No.
Mustafi has been needed in Mertesacker’s absence (and Gabriel for a while also). But although he has many strong qualities, he also seems a little weak in the air and prone to mistakes.
Xhaka. Is he a significant upgrade on Coquelin and Elneny at the base of midfield? I would suggest not. Not desperately quick and prone to indiscipline (albeit with a nice shot and pass). In the box-to-box role he has Cazorla and Ramsey ahead of him (and possibly eventually Wilshere).
Lucas. Haven’t seen much. Clearly still adjusting to the English game – and now injured. A bit of a concern for a 28 year old.
But, it is too early to make definitive judgements. Things should become clear in the new year.
The acquisition who really has impressed me is Holding. I think he has huge potential and would not be surprised if he forced his way into the first team in the next 12-18 months. I suspect we might see Gabriel sold fairly soon (if the rumours I hear about his dissolute lifestyle are true) and I’m not sure that Chambers will ever make it with us.
But, having a larger squad has certainly helped given the usual injuries.
So, to answer Tony’s question, no I don’t think we need to buy anyone. There may be one or two youngsters and, on the offchance a superstar becomes available, just conceivably.
OT: Amused myself by watching the last half an hour of our neighbours’ futile efforts at Wembley. Shouldn’t be overconfident but they weren’t looking good. And can someone tell me what all the fuss is about Alli? I have watched him a number of times for the Totts and England and he really hasn’t stood out at all (and that’s being kind). I know he is young and has plenty of potential – but unclear where the hype has come from?
@Pete.
You had it nicely said there. But I doubt if Gabriel will be sold next summer by Le Prof for another new CB even if we won the title this season. Not to talk of selling him in January, he won’t. And I think Chambers will eventually makes it at Arsenal as a top centreback. Time is still on his side as he has already shown he has the potential to be a top CB for Arsenal I would think.
Let’s not forget Arsenal are no longer a selling club that they were forced to be some years back. And beside this, Le Prof don’t gives up easily on a player he believes will eventually comes good for Arsenal. He gives them time to prove if he’s right or wrong.
I doubted my claim, hence I checked Totts results only for me to see they’ve even lost 2-1 to Liverpool in the ELC last 8. So there unbeaten run to the season in all competitions had gone up in smoke before their Bayer Ucl game of yesterday.
Yes we have always been shot in the middle of the park.. We need one of those selective visioned whistle blowing controllers of the game. It is time Kronke got his book out & picked one select official for Arsenal. Wenger doesn’t care about them any more. He will win despite them. I feel however it would be great to have our own official miked up & on the PA commenting on the PGMO bod in the middle. It will shake a few trees & many coconuts will fall to the ground amidst the laughter from the gooners.
OT: November being a tough month for Arsenal
No questions, eh?
While the league is sort of scheduled such that there is a game every week, we sort of expect that the number of games per month should be something like 4 as there are about 4 weeks in a month (and I am assuming a month is 30 days).
Over the reign of Wenger, the minimum number of games per month is 3.3 (which probably means 3) and the maximum is 5.8 (which probably means 6). The number of data points is 236, which is reasonable, and the average is 4.1 games per month.
The X axis is days since Jan 1 of the year the season starts, and the season typically begins about Day 230. The peak is at day 352, which would be around Dec 15. You can see features in the data which corresponds to common tendencies in the international schedule, and there is a rising tail to the data, due to replays needed for postponements due to weather and needing to get the season finished before the last day. The month of November is one which tends to have more games at the end of the month, than at the beginning.
File gpm.gif to be emailed to Tony.
I’ve produced a second plot, which starts with the above data, and adds to it the average number of points per game, over a 4 game stretch (so NOT a month in general). This second data is bounded 0 to 3, so it lies entirely below the games per 30 days data.
You can see a whole in the observation of 3 points per game expectation in a few places, but most noticeably in the November and possibly some of October, and possibly some beginning of December. I wouldn’t want to say it is statistically significant. You do see a horizontal hole in the data, as it is not possible to get 11 points from 4 games.
File gpmppg.gif also emailed to Tony.
So, I will ask again. What questions do you have of the data?
Do we actually need to buy anyone this January?
Yes please! Lionel Messi will do nicely as a backup for Sanchez and what a super sub he would make! I’m sure even Giroud would agree!
Q. Do we actually need to buy anyone this January?
A. NO .
Am looking forward to a very quiet January transfer window .
Yawn.
A tree that wants to touch the sky must extend its roots into the earth.
The more it wants to rise upwards , the more it has to grow downwards .
So to rise in life , we must be down to earth , humble and grateful.
We won’t need to buy anyone this January – unless we suffer real misfortune with injuries.
We may buy a left-back and some younguns for the future. The former depends on all sorts of ifs, buts and maybes: is Monreal’s contract up next summer? If so, will the club renew or does the player want to renew?
But I can see us selling Debuchy and Sanogo. There have been many rumours that they are, or have been, injured, but I find it odd that there’s been no official news.
Summer 2017 will be more interesting. What will happen to Mertesacker and Cazorla? What will happen to all the players out on loan? Six or so of our young players (including Toral and Akpom) will no longer classify as U21, so I can see us losing more than a few. Of course, the biggest question will be does Wenger stay?
The curse of Wembley strikes again. A record 85000 thousand supporters watched Spurs lose 0-1. I’m sure there’s some irony in there.
Personally, I don’t see a reason to sign a new player in January. We have enough players and with players such as Welbeck and Lucas back in 2017, we should have enough reinforcements from our treatment room solely. We got all we had needed this summer including a central defender and a new starting central forward. Now we have over 50% of games in which we have scored three or more goals with fewer chances created.
Mind you, I had seen a need for a new midfielder before last January so it’s not that I would be silent about it. I just feel we have enough squad depth to deal with the season.
I do think we should. That is:
If a player becomes available that fits Arsenal we should.
This means that some player(s) may well be on their last hours at Arsenal.
I cant tell you who, but AW must know who is at the bottom of the list due to performance/injuries/mindset/etc. and it will be time for them to start looking elsewhere.
See, Arsenal needs to continue improving.
Unlike Pete, i think that Arsenal has improved, but that does not mean we stop improving where we can.
Gord
Your hard work in creating the database should not go unrecognised. There should be many questions and theories it could help test. Can you see sequences that could be drilled into e.g. did our world really fall apart after that league cup loss to birmingham (sorry to reopen old wounds)? Were there events that kick started a winning streak. How do we perform against certain teams or category of team.
Does the festive fixture madness feature (apart from having such a high frequency of games)?
We have to recognise that low sample sizes can influence the appearance of results (especially if applying %) and even whether they are justified and it is good that you do comment on this.
Looking at November this year 3 epl games Spurs Manu and Bmouth what average would we expect (apart from 3). 2 cat A games with PSG and internationals thrown in. December with 5 EPL games would be different? At least it’s Stoke at home.
I’m with Pete really. The additions who’ve come in have helped more in kicking the existing players on, rather than any great skill advantage. The only reason I can see Mustafi here is because Gabriel is gone in the summer, and Le Boss made an early move. That apart, while I hope he improves.. he runs forward far too much ala David Luiz. Its early though, and there’s also promise shown – on and off the field (he comes across as a well spoken young man in his interviews).
Xhaka honestly has had a fairly poor start. Again I hope he improves and there’s certainly time for that, but hey he’s got 3 goals already, so the AAA and company will be wising up and demand a new shiny DM in January. Coz its definitely not Granit :D. Mo Elneny should play more, there’s no question.
Lucas is a pure Plan B option. Much like Reyes all those years back. He’ll be a key part of the puzzle, score a few goals and leave in 2 years. May win us the title. I’ll say it now.
Rob Holding. Easily the best of the lot. Hope he keeps developing.
p.s Am I the only one who finds the Ozil goal hype a little over the top? Everyone.. Arsenal.com included have forgotten that Giroud bust a gut to get back and tackle, Elneny spotted the run with an inch perfect pass and only then could Ozil work his magic. Don’t get me wrong, it’s nice to watch.. but really Mo and Olly should be taking a lion’s share of the credit. The overall defending was quite poor, if you ask me for the goal. Skilful yes, lovely to watch yes… but surely the credit should be shared more.
And no. Please dont buy more. Please. I’m disgruntled as I am this season with the overbuying. Yes. I do think so. I could be wrong, but I do think so.
Tony
WHAT HAPPENED TO THE aaa/won?
Don’t you just love it to see our neighbours on the way to reaching at best Europa League qualifiers ?!?
And suddendly they seem to have a full sick bay.
For the record, we are one of only 6 teams that have qualified after 4 games.
Yet still you read about how reaching last 16 CL is not a trophy and that we have so many injured playaers and the worst injury record.
As Tony so damningly made the point in his piece, the press just won’t do anyting else than report ‘against’ Arsenal. Their bias is at least as bad as the referees’ usual bias.
And when the tinies will fail to qualify, no one will scream for the manager to be sacked. I can bet you we’ll read stuff like : this is a learning process, it takes time to make champions, we had too many injured players, the grass in Europe is not as good as in London, it was the russian hackers fault….etc etc….
Yet, I just watch that goal over and over again. And am happy with Arsenal.
Chris you are right re
‘I can bet you we’ll read stuff like……’
On Talksport this morning I have heard no Spurs criticism to speak of, more like sympathy for their plight. I think we took more stick for only just scraping through against Ludogorets (according to the experts we should have coasted through) the previous day than Spurs have so far for last nights shambles.
Off topic but just came across this http://www.parliament.uk/edm/2003-04/1044
January will be the time those journalists needs to be replaced themselves.
Spurs can’t buy a goal at the moment – they won’t trouble us this weekend and I’m as upbeat about a derby as I’ve been in years. Expecting a shutout and a few goals from us, possibly giving them a pasting. 3 or 4 nil is entirely possible given how we’re both playing. No new personnel required in January – this squad has depth and balance and simply needs to keep it’s composure and confidence through the season for great things to happen.
I think it’s highly unlikely Arsenal will buy anymore players this January as AW did say he was looking at buying only 3 players and he bought the 3. I think Arsenal might look at just getting a left back as a replacement for Monreal or backup to Gibbs in next summer transfer window.
I think BFG will leave when his contract is finish and most likely head back to the Bundesliga, Kos probably have another season or two before he probably leave due to age. I think Holding is Kos future replacement and Mustafi is BFG. Chambers and Gabriel will be rotational players.
I think Santi due to his age, Arsenal will most likely offer him a per season base contract instead of a long term one. If Santi reject this offer, I think he will head back to La Liga and Ramsay or El Neny will fill his spot.
If reports are correct that Ozil wants assurances that AW will stay as manager before signing new long term contracts, then this could be an indication that AW planning to step down after this season.
Robido
Your questions by and large seem to be too specific. But I’ll see.
I sent images to Tony, Tony wants me to write something. A something with 2 images probably isn’t enough, I have to play with the data a little to see what can and can’t come out.
Okay, since the beginning of the 1996/97 season, there are 706 sets of data covering average points per game over 4 contiguous games (in my data set). There is this ghost of a hole in about November. Is it real?
The average points per game is 1.994, and the median points per game is 2. Which means that by and large; the average behavior of Arsenal has the same result as never losing but only alternating between wins and ties. Which works out to 76 points per season. Standard deviation is 0.615.
That the median and mean are so close to each other, likely means that in the vicinity of the mean, the distribution of points is symmetric. Whether one looks at the “distance” that 25 and 75 percentile are from the median, or 5 and 95 percentile; they only differ by 1 (in 706 points). This distribution is very symmetric. The 5 percentile point is 1 point per game (average over 5 games), so less than 5 percent of the data is worse than 4 consecutive ties (there are other ways to get 4 points from 4 games). There is a slight feature in the data, as it isn’t possible to get 11 points in 4 games, which happens to make the 95 percentile point 3 points per game.
As unstructured data, it seems very unlikely that November is difficult month for Arsenal. But as the data is actually a time series, we can do more things.
Since nearly all teams purchase players and players may be sold or retire, there may be a reason for a linear trend to exist in the points per game data. But, with games it is reasonable to expect that teams play better.
It is a little more complicated than that, as we also have this transfer window mid-season. So there should be sort of a saw tooth affect.
If a team has players out on international duty over the summer, they may pick up injuries or become fatigued, and hence not be in a position to play as well as they could at the beginning of a season. So, we might see a short linear trend at the beginning of the season on top of a linear trend for about half the season, then a drop when new players come in, and then another linear trend.
One way to minimise the mid-season drop, is to minimise the number of new players brought in mid-season.
Okay, we have a time series. Or, since these are seasons of football, we have many time series, it just happens that one follows another.
But we are kind of expecting that a season consists of a ramp function, a drop and another ramp function. Not only that, but we expect the amount gained over the second ramp is bigger than the drop, so that the team performs better at the end of the season than say at the end of September (6 weeks in). We may also see a sub (steeper) ramp to the first few weeks of the season, due to obligations to the FIFA international system.
The first thing we are going to assume, is that this is a global phenomenon, it happens each and every year. So, I can pool together all 20+ years of data. What I want to do, is fit a straight line to this data. If I am fitting by least squares, I know in advance that the line will pass through the point (mean(X), mean(Y)); and hence if I subtract off mean(X) from all X values and mean(Y) from all Y values, my fitted line passes through (0,0). And the statistical test I need to first apply, is to test if the slope found is statistically different from 0.
If the slope is not statistically different from 0, that means that in terms of all the data, we are expecting that the Y values are all the same (they are independent of X).
If the slope is statistically different from 0, we proceed to subtract off the linear trend from all values, so that our new Y value is now independent of X.
We then fit a straight line to each year of data (except for 2016/17, where we have less than 1 season of data). We have 2 things to look for, does our fitted line pass (approximately) through (0,0) and does it have a slope not statistically different from 0. If these are both true, it means that the data from that year, at least superficially resemble the data in whole.
If the fitted line does not pass close enough to (0,0), it may be due to calendar effects: is it a leap year, or did that year begin significantly before or after the average beginning. Another thing that might show up here, is games being postponed or called off due to weather. Those kind of situations would lead one to shift the data in X. And at that point a person really should begin the whole analysis anew.
If it doesn’t look like a calendar effect, and the differences are significant, we subtract off a new linear trend just for that year.
And we do that, for all 20 years of data.
Now we partition the data into data after the mid-season transfer window closes, data in the first 6 weeks of the season (6 weeks is an assumption on my part, has to be verified or adjusted by the data), and what is in the middle.
We now have 60 sets of data, all smaller than 38 data points. And we do the same straight lines fits and adjustments to all of them. Or maybe we look at the global 20 year trend first, subtract that off and then do the 20years times 3 of all the years.
At the end of all that, we can suggest:
_1 in what years there is an international service effect, and how big
_2 in what years there is a mid-season drop off, and how big
_3 does the team improve across the year after allowing for the drop offs? Is it the same before and after mid-season transfer window (if we bought Ballotelli in one mid-season, it wouldn’t have been).
We aren’t done. But, we have a pile of linear trends we can remove from the data. In an ideal world, if we were to glance at the data now, it looks like random dots in a band of data with no obvious trends. But, we have changes in fixture density at Christmas (always), and probably at the end of the season in select years due to weather effects (industrial action might be another contributor).
My guess is that changes in points per game due to changes in fixture density is not linear. I suspect a curve, and my inclination would be to fit a parabola to regions. I have 2 parameters, the symmetry line of the parabola, and some measure of its curvature. Actually 3, the extent over which is “quadratic effect” manifests itself.
After all that, I think we may be able to look at the (adjusted) data and ask, is there a November effect?
Now I ask you, have any sports journalists done this? Hmmm?
They aren’t reporting news, they are writing STORIES. Works of fiction.
===
There is a lot of details here, and I have other things outside of football that need to get done. So this might be my contribution to Arsenal this season. And it won’t be done tomorrow. Sorry.