GW2 and GW3: both sides held

Two gameweeks since the last note and I’ve let both 15s run as they were. No transfers in either, no chips, free transfers rolling. I did fiddle with the Haaland-in side inside the squad though, I gave Haaland the armband for GW3 instead of Bruno and went 4-4-2 with Diop in and Thiago on the bench. Verbruggen is in goal on both sides now, with Dubravka sitting behind him on the Haaland-in one.

Where that leaves things after three weeks is Haaland-in 184, no-Haaland 181. Bear in mind the T-48h packet had the no-Haaland robust arm 21 points ahead on its six-gameweek score (255.88 against 234.32), so seeing the Haaland-in side edge in front over the last two weeks is interesting, even if I wouldn’t read much into it yet.

The retain decision

Going into the GW2 deadline the 27 Aug hold briefing had set out what would need to happen for it to move the Haaland-in side, and on the Friday none of it did. Gibbs-White was in Forest’s XI, Kinsky kept the gloves at Spurs, and there wasn’t a fresher packet to rescore against. So I banked the free transfer, which is what the briefing said to do.

Same again going into GW3. Two free transfers in hand, nothing on the official injury feed touching the XI, packet still degraded. The agent has been pretty consistent that transfers are there for positions that have been falsified rather than for chasing whatever scored last week, and I’ve gone along with it. Xhaka’s 9 on the bench in GW1 is the sort of thing that would normally have me reaching for a move, and he’s gone 6 and 3 since, so fine.

The bits I did change for GW3 were free. Haaland took the captaincy off Bruno and Diop came into a back four with Thiago dropping out. I’m treating the 15 as the thing the loop owns and the XI and captain as mine to play with week to week, at least for now.

So that’s three free transfers banked into GW4 (deadline Saturday 12 Sep, 13:30 UK). I’m not fully sure whether that’s discipline or just not getting round to it, more on that further down.

GW2

Average 81. Highest 161. Haaland-in 90, no-Haaland 91.

Bruno got 3 goals and an assist at home to Ipswich, 23 raw and 46 with the armband, and that pretty much was the week for both sides. It’s also the captain call the packet had been making since 12 Aug (Bruno 6.37 EP over Haaland 5.74), so nice to see it land. Haaland got 13 at Palace, so having it the other way round would have cost 10.

Haaland-in 90

Bench: Dubravka 0, Xhaka 6, Diop 2, van Ewijk 2. All 11 played, no autosub.

Gibbs-White, who was the knee flag in the Friday hold, played 90 minutes against Liverpool and came away with a goal and an assist.

No-Haaland 91

Senesi didn’t play so Shaw came off the bench for him. Rogers 5 unused, Verbruggen 0, Obi 0 minutes.

Seven different players across the two XIs and they finished a point apart. Bruno and João Pedro were in both and did 32 of each side’s total, so the other nine on each side did 58 and 59 respectively, which I wasn’t expecting.

GW3

Average 51. Highest 119. Haaland-in 65, no-Haaland 42.

First week the two sides had different captains. I went Haaland C on the Haaland-in side and left Bruno C on the other. Bruno got 2 at Everton, Haaland got 9 at home to Coventry with a goal, a clean sheet and three bonus, so that’s 14 off the captain slot against 4 and the switch was worth 7 over sticking with Bruno. Worth saying the packet numbers still had Bruno, this was me going over the top of them.

Haaland-in 65 (4-4-2)

Bench: Dubravka 0, van Ewijk 3, Xhaka 3, Thiago 2. All 11 played, no autosub.

Mitchell scored twice at Fulham. The cheap defence the 12 Aug change bought (Van Hecke, Mitchell, Shaw) did 27 between them this week, having done 3 in GW1 and 3 in GW2, so that’s how that goes. Diop in for Thiago was worth one point either way.

No-Haaland 42

Two autosubs this time. Senesi didn’t play again so Shaw came in, and Beto has gone to Fiorentina permanently and is now unavailable, so Rogers came in for him. Verbruggen 3 unused, Obi 0 minutes.

The Arsenal three that carried GW1 and GW2 (Raya, Gabriel, Rice) got 10 between them against Chelsea. If the autosubs hadn’t kicked in this side scores 32.

Three-week ledger

Both sides are basically on the season average, a couple of points either side of it, so neither is doing anything special against the field. The Haaland-in side has made up the 19 it was behind after GW1 though, and Haaland at 15.5 has 24 points to Bruno’s 27 at 12.0, which is roughly where I’d expect that argument to be sitting.

Correction from the last post. I had GW1 as average 36, highest 114. Those were the provisional numbers from the Saturday night, the API now has GW1 at 50 / 131 and that’s what I’ve used in the table.

Falsifiers, checked

The 27 Aug hold listed five things that would change the Haaland-in side. Here’s how they did:

One out of five. Harry Wilson is now a falsified position by the agent’s own definition, so really the loop should be putting a transfer against him rather than me banking a fourth, and that’s the bit I mean about discipline versus not getting round to it.

There’s no equivalent list for the no-Haaland side because that one came straight out of the optimiser rather than the strategy agent, which I should probably fix. Two of its 11 have stopped playing (Senesi, Beto) and it’s been living off the bench for two weeks. Beto going to Fiorentina takes that decision out of my hands at least.

What I take from the models so far

The captain pick is the one place I’ve used the packet numbers directly. Bruno over Haaland was a wash in GW1 (both got 2), gained 10 in GW2, and then in GW3 I went against it and captained Haaland, which gained 7. So following the packet is +10 so far and the one time I overrode it, I got 7 out of it. Obviously nothing to read into at three weeks but it is the first override I’ve made so I want to keep a note of it as we go.

The hold-unless-falsified rule has done what I wanted, which is keep the Haaland-in 15 steady through a 29 and then a 90 without me reacting to either. Where it’s weaker is that the falsifier list only exists for the advisory side and only covers what the agent could see at T-24h. Wilson’s minutes drifted down over three weeks but the trigger was written as a single-week collapse, so it took a while to formally trip.

And the six-gameweek comparator had the no-Haaland arm 21 points clear. Three weeks in it isn’t. Far too small a sample to say anything about the scorer but this is the reason I put both in.

Into GW4

Three free transfers on each side. No chips used.

Haaland-in: Wilson is the falsified position. Everyone else in the XI played 90 in GW3.

No-Haaland: Beto has to go. Senesi has two blanks in a row so that’s a minutes question too.

Captain: the packet has been Bruno every week and I’ve already moved off it once. GW4 is United v City at Old Trafford so it’s Bruno or Haaland against each other’s defence, which doesn’t make it easier. I’ll wait for a fresher packet before deciding whether GW3 was noise or the start of a Haaland C case.

Both 15s are as listed in the GW1 post, with Verbruggen and Dubravka as the Haaland-in goalkeeper pair.