John Spencer v. Gavin Wilkinson

At this juncture, I thought it’d be interesting to see how Gavin Wilkinson’s team is performing against John Spencer’s. It’s a bit earlier than I’d have expected to pass judgement, but so much has happened.

The data I’m looking at was averaged across the last 8 games. Three of which Wilkinson has taken charge – home to LA Galaxy and away to Chivas and Dallas. The other 5 were under John Spencer’s reign – home to Seattle and San Jose, away to Colorado, Real Salt Lake and LA Galaxy. I’ve averaged out all of the statistics I talk about.

John Spencer pretty much stuck with the 4-4-2 diamond formation. Two up-front was a favoured deployment of Spencer’s, the Boyd/Mwanga partnership was one he particularly wanted to blossom. The only time he diverged from that was in his last game, the 3-0 loss to Real Salt Lake. More on that in a second.

He was only able to use Kimura at Right-Back once, so with Jack Jewsbury deputising there, Diego Chara sat at the bottom of the diamond in the defensive midfielder role. Part of my criticism of Spencer was the inability to spot that playing 2 wide-men in a diamond is a fatal flaw. Songo’o and Alhassan would sit either side of protégé Darlington Nagbe, who Spencer had attempted to convert to a more central role behind the front two. My main complaint of this setup was that pressure that put on Chara to protect his back four. The most effective midfield setup actually came against San Jose Earthquakes, where Eric Alexander played on the left prong of the diamond, providing balance and support to both Chara and Nagbe, leaving Franck Songo’o to have his best game in a Timbers shirt. This all went to pot against Real Salt Lake when Spencer unveiled his 4-3-3/4-5-1/Kitchen-sink hybrid. Less said about that the better.


Jewsbury               Mosquera                           Brunner                  Smith


Alhassan                                       Songo’o


Boyd                        Mwanga

John Spencer Typical XI

Wilkinson has built on what Spencer attempted, and for the first 2 games of his tenure, played Kris Boyd as a lone striker. At Jeld-Wen, Boyd showed exactly what he could do in this position against LA Galaxy. Having a hand in all 3 goals with some excellent wide service from Khalif Alhassan. Chivas, however, was another exercise in how not to deploy the Scot. He was chucked on the bench against FC Dallas for “climate” reasons.

The midfield setup has been… interesting… under Wilkinson. Against LA and Chivas, we lined up with Jewsbury and Palmer (LA)/Chara (Chivas) alongside each other in the DM positions. That midfield was completed with Eric Alexander out wide-left, Alhassan in his familiar right-wing role and Darlington Nagbe supposedly behind Boyd, but in reality, who knows where he was. The issue I had with this is that Jewsbury does nothing of note when he plays alongside another DM, and Eric Alexander isn’t a left-winger. Well, against Dallas we lined up in a 4-4-2 with Alexander and Chara in the middle (all prayers answered), Songo’o on the left, and Alhassan on the right. A midfield that got completely ran through. Shows me what I know.


Kimura                   Mosquera                          Futty                           Smith

Chara              Jewsbury

Alhassan                    Nagbe                 Alexander


Gavin Wilkinson Typical XI

The personnel hasn’t changed much. The only differences in defence is that Wilkinson has been able to choose Hanyer Mosquera regularly. Spencer was stuck with no right-back, and a Horst/Futty partnership that was destroyed on the road twice. Nothing new then. In terms of the midfield, I advocated the way Wilkinson shaped us against LA and Chivas. I’d argue about who filled what role (Chara & Jewsbury never works), but on the whole, I was happier with the intent. However, what resonates between the two managers, is the willingness to abandon system when it doesn’t go quite to plan. Yeah, we lost to Chivas, but I don’t particularly think Wilkinson’s system was to blame. I also think Spencer was close to getting his own system right. The formation looked solid against the Quakes, although there seemed to be a lack of confidence in individual role.

Okay, down to the nitty-gritty. Statistical comparison.

Using Spencer’s last game against RSL as an example, I bemoaned the fact that Kyle Beckerman had completed more passes (80) alone than Chara, Jewsbury, Alexander and Palmer combined (75). It was something that had to improve. Looking at GW’s first 3 games – it actually has.

Against RSL, we’d only managed a pass success rate of 71.2% from 337 attempts. We were completely eclipsed by RSL’s 83.2% success off 565 attempts. Even in our win against San Jose at Jeld-Wen, we only managed to string together 254 out of 397 passes – an awful rate of 64%. San Jose weren’t much better mind, but they still attempted 128 more passes than we did.

Our first post-Spencer game saw a hugely positive upswing in our passing numbers. Where just 1 week before we’d attempted a paltry337 passes, against LA, we attempted a mammoth 630 passes. 323 more than we did against RSL, almost double. On top of that, we’d completed 509 – a pass success rate of 80.8%. Even against Dallas, we hit an awesome success rate of 83.7%.

Personally, I attribute that to a proper definition of roles. Spencer’s ideas were fraying at the seams, ultimately unravelling against Beckerman, Saborio et al. Against RSL, I made up the below heatmap .gif showing that lack of understanding. 

Compare it to the LA game, and you’ll see a real tightening up of roles, a real definition and purpose to the individual. Yes, the “spread” of activity remains similar in the Dallas game, but the propensity to work for that role – the lack any real pattern and idea disappears against Dallas.  The unfortunate thing for Gavin Wilkinson, is that it seems that the more time spent with his players, the less effective his methods of communicating their job descriptions become. See the below, versus FC Dallas:

Once again, the headless chickens flood the middle of the park.

Another improvement is that in the number of crosses we’re now attempting. On average, Spencer’s Portland would attempt 11 crosses and 3 of which would be successful. Wilkinson’s Portland attempt 19, again with only 3 successful. So whilst the end result has not improved, the effort and focus has. I’d attribute part of this to the inclusion of Kimura – a player Spencer only used once. Whilst I have serious reservations about Kimura’s defensive ability, he does give Kalif Alhassan a lot more support on the right-hand side in attack. Hopefully, this is a combination persevered with. It will, however, only work if Kalif understands the importance of wingers tracking back.

Possession-wise, again the stats seem to swing in favour of Gavin Wilkinson. Where in Spencer’s last 5 games our share of possession averaged at 43.4%, Wilkinson’s works out at 51.4%. We also lose the ball in possession less than John Spencer’s Timbers. On average, we lost the ball 110 times against LA, Chivas and Dallas. Spenny’s team would lose the ball 135 times. In saying that, the behaviour of our opponents has changed also. Team’s against GW’s Portland lose the ball less too – 107 times on average. JS’s opponents lost the ball 138 times. It’s almost as if the games Spencer was involved with were a bit more lively? More tackles, more enforced errors, all-in-all tighter games? The statistics suggest so – on average there were 90 “duels” a game versus 81 in the Wilkinson games. There were also more tackles – Portland winning 13 tackles on average a game to their opponents 15 under Spencer, and winning only 9 tackles to their opponents 11 under GW. Against Wilkinson, our opponents make 9 successful dribbles a match, to the 6 made against Spencer.

We’re obviously talking about slim margins here, but we know what that means in reality. In that famous scene from Any Given Sunday, Pacino said that Football (of a different kind) was a game of inches. Well, so is our Football, except it’s not about gaining inches, it’s about the difference an inch can make. The findings in possession, to me at least, would suggest that Spencer was more likely to deliver a Pacino-esque speech to his players – in other words, they seem to fight for that inch. They’d stop just 3 more dribbles from being successful. They’d win just 9 more duels. They’d attempt more tackles. They’d force more errors from their opponents.

The statistics I pulled together in regards to the defence didn’t show much of a difference between the two, to be honest. I think the defence has been a big issue for the whole season, and it continues to be so. You look around the league, and we don’t have a good enough center back – simple. We don’t have that experience required – a Jay DeMerit, a Rafa Marquez, a Danny Califf. Regardless as to whether you rate them or not, a player with experience, positional sense and most of all – a voice.

Both Spencer and Wilkinson have had trouble finding two from Horst, Mosquera and Futty, in the absence of Brunner. Horst is too green, Mosquera hasn’t had enough MLS minutes under his belt and Futty is just too prone to a lapse in concentration. As Manager and GM, they’re both just as responsible for a lack of a leader at the back as the other is.

Statistically, I don’t have a big enough sample of games to look into whether or not Wilkinson’s going to get the best out of our current strikeforce or not. Currently, the only game where we saw what Kris Boyd can do was against LA at home. To be honest, we never saw him play that well for John Spencer either. I’m interested to see how Boyd performs against Chivas at Jeld-Wen, when he’s in a more comfortable, Northern climate.

What I can say, is that we still have a problem in supply to him. It actually runs deeper than just a “supply” issue, it’s more an expectation of what Kris Boyd can do for us. I’ll not rehash my belief that Boyd is an opportunistic striker, and has to be supported as such, any more than I have already. What I will add to that argument is that far too much falls on his shoulders. If you’re going to play him up top by himself, then you sure as fuck need Darlington Nagbe to understand what his role is. Nagbe’s a really, really tidy player with a lot of potential. But he needs reigning in, and he needs to know what is expected of him. If it were me, I’d tell him to play up front with Boyd, but ask him to keep dropping deeper/wider and support his midfield. At the moment, he has no clue where to stand, never mind what to do with the ball. I’ve also been disappointed as much with the deployment as I have been the performance of Danny Mwanga and Mike Fucito. The latter especially.

Ultimately, neither Spencer nor Wilkinson decided what their plans were for Alhassan, Songo’o, Fucito, Nagbe, Mwanga and Boyd. They’ve been thrown on the pitch and expected to gel. There’s next to no dynamic between any of them, and neither Spencer or Wilkinson know the best attacking combination to put out on the pitch.

Ultimately, in my mind, it boils down to two things. The second is in reply to what I’ll say first.

Under Gavin Wilkinson we have scored 3 and conceded 11. As a rounded up, average score, that’s a 4-1 loss. In Spencer’s last 5 games, we scored 4 and conceded 9, giving an average result of 2-1 loss. Not good for either manager, but one decidedly better than the other.

Where I take umbrage to Wilkinson is in his handling of the players. As pointed out in a comment on my last post (I can’t take credit for coming to this conclusion), the man who stands on the soapbox bemoaning the lack of ability, the lack of fight in the current roster, is the man who decided that they were of good enough quality to get the job done for his predecessor. Gavin Wilkinson had arguably a bigger hand in getting these players onto the Portland roster, now he’s selling them down the river to make himself come out of the shit-pile smelling of roses. In doing that, he’s doing himself out of a job. In singling out individual performances, he’s inadvertently admitted the players HE brought to Portland are just not good enough to compete in the MLS.

Yes, Spencer had to go. But so does Gavin Wilkinson. Not just from the interim First Team Coach job that he currently holds, but his day job as General Manager also. Paulson expected Wilkinson to tide us over til the end of the season, but in reality, pulling him downstairs was never going to be any different to keeping Spencer on in his role. Both men were responsible for the playing and coaching staff at this club, all Paulson has done is ripped the heart out from it. 

Agree? Disagree? What positions do we need to strengthen in? More importantly – who do we strengthen with?!


6 responses to “John Spencer v. Gavin Wilkinson

  1. First of all love the website, it’s a go to site for me. Aaaaand holy crap those heat maps are bad, like these absolutely represent what I saw and hid my child’s eyes from on the field.

  2. Let’s assume that Merritt knows what he’s doing (at least, I hope he does). He hasn’t been around football all that long. He is the new kid. He doesn’t have a lot of friends in the sport. How does he look for new management without an old boy’s network? His father is a big wig, but that guy is having lunch with Bob Rubin not Bob Bradley.

    Even assuming there is (at least, I hope there is) some old hands puttering around the club house. What are are Merritt’s options exactly?

    1) He’s got a high profile coach who was a high profile player. Talent or no talent John Spencer was somebody you knew in the world, right?

    2) He’s got a local boy GM who’s been with the club 12 years who he is realizing doesn’t cut it (at least, I hope he realizes).

    If he fires GW, nobody cares. Externally, nothing changes. If he fires JS, people stand up. At least, people notice. Maybe they call him.

    My point is this: John & Gavin were fired on the same day.

  3. Not so sure. Gavin is a survivor with this organization. He has – in my opinion, already made enough poor decisions to get him canned in many other outfits. His handling of Keita, and then Pore in USL. His ridiculous tanking at the end of the 2009 season (yeah, he lost Cam Knowles, but if your defense is so dependent on one player that it collapses when he goes out you’ve got problems, there, head coach…) His hoarder-like stockpiling of forwards as GM without ever really trying to take a whack at the problems at midfield and in the back…

    So I’d like to think that Merritt is currently scouring the soccer world for a head coach who can also serve as the GM, a guy with top-flight experience, a matchday savant who is also a good team-builder and has connections outside the local community…what I’m afraid of is that we’ll get somebody in as coach who will be stuck with Gavin as GM making the player signings and continuing the “just-good-enough-for-USL” standard we’ve maintained up until now…

  4. Decent analysis and good reading, but I’m going to take issue with the passing section. I’m not sure it’s accurate to say that passing has improved under GW. I ran through a lot of these OPTA numbers myself right before Spencer got sacked, and the RSL game sticks out as a horrible departure from a norm that is itself pretty bad. Chalk that up to looking at a team playing what they know is a last-ditch effort from Spencer. Chara had a season average through RSL of 83% pass completion on 44 passes/game. Taking him as a symbol of the team as a whole (when Chara plays well, the team plays better), I think it’s premature to say that pass accuracy has improved. The *quality* of passing has improved imho, but that’s a product of taking Palmer off the field.

    • Hey, thanks for the reply.

      In my opinion, in the immediate, the passing has improved. It’s one component of the team that has. However, we’re not creating the right sort of chances, and we’re certainly not converting as many as we were/should be. I also can only look at the 3 games (this was before last night’s 1-0 loss to Chivas) GW has taken charge.

      John Spencer’s last 5 games in charge (on average):
      Portland 0-3 RSL
      240 successful passes, 97 unsuccessful, 71.2% success rate

      Portland 2-1 San Jose
      254, 143, 64%

      Portland 0-3 Colorado
      303, 107, 73.9%

      Portland 2-1 Seattle
      214, 113, 65.4%

      Portland 0-1 LA
      368, 104, 78%

      Portland Average over 5 games – 276 successful passes, 113 unsuccessful, 70.5% success rate.

      Gavin Wilkinson’s first 3 games in charge:
      Portland 0-5 FC Dallas
      338 successful passes, 66 unsuccessful, 83.7%

      Portland 0-1 Chivas USA
      438, 88, 83.3%

      Portland 3-5 LA
      509, 121, 80.8%

      Portland Average over 3 games – 428 successful passes, 92 unsuccessful, 82.6% success rate.

      Our pass success rate has improved in the immediate. As my article said, this doesn’t count for an overall improvement in the quality of passing, chance creation or general play. The results obviously show that.

  5. Pingback: Toronto FC v Portland Timbers – Preview | #RCTID·


Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s