GW32 -- Preview

TLDR:

• Both models captain Bruno Fernandes despite only 1/11 starting XI agreement
• v2 model projects 2.9 points higher (53.8 vs 50.9) by trusting baseline projections more
• Both trigger Wildcard in GW32 with 9-10 transfers, then BB/FH in consecutive weeks
• Major divergence on Haaland: v1 downgrades him 1.09 points while v2 keeps baseline
• v1’s z-score adjustments systematically downgrade players with 0.00 elite ownership

Squad Picks (GW32)

The models show stark disagreement with only Bruno Fernandes matching in the starting XI. v2 favors Arsenal assets heavily with Gabriel (5.20 xP), J.Timber (4.60), and Gomez (4.15), while v1 spreads risk across Newcastle’s Dúbravka and Brighton’s Van Hecke. Both include premium midfielders Rice and Gordon, but v1 opts for Wilson over v2’s Gomez in the final midfield slot. The 2.9-point projection gap suggests v1’s elite ownership adjustments are overly conservative for this gameweek.

Slot	v1_Player	v1_xP	v2_Player	v2_xP	Match
XI	Dúbravka	3.58	Verbruggen	3.66	~
XI	Thiaw	4.58	Gabriel	5.20	~
XI	Gabriel	4.44	J.Timber	4.60	DIFF
XI	Van Hecke	4.29	Thiaw	4.29	DIFF
XI	B.Fernandes (C)	7.39	B.Fernandes (C)	7.39	=
XI	Gordon	4.60	Rice	5.02	~
XI	Wilson	4.36	Gordon	4.43	DIFF
XI	Rice	4.34	Groß	4.40	~
XI	Groß	3.77	Gomez	4.15	DIFF
XI	Thiago (V)	5.18	Haaland (V)	5.45	~
XI	Haaland	4.36	Thiago	5.18	~
B0	Verbruggen	3.45	Dúbravka	3.58	~
B2	Guéhi	3.03	Guéhi	3.28	=
B3	O’Reilly	2.92	O’Reilly	3.27	=
B1	Calvert-Lewin	3.20	Calvert-Lewin	3.68	=

Transfer Plan

Both models execute similar Wildcard strategies in GW32, making 9-10 transfers to completely reshape squads for the run-in. The chip sequencing is identical: WC32, BB33, FH34, preserving flexibility for the crucial double gameweeks. Key differences emerge in GW36 transfer plans where v1 makes a single Groß to Sarr move while v2 plans a double transfer replacing Gordon and Gomez with Wilson and Sarr. This suggests v2 anticipates needing more tactical adjustments as fixtures evolve.

GW	v1_Transfers	v1_Chip	v2_Transfers	v2_Chip	Match
32	Darlow, Raya, Andersen, Virgil, Enzo, Ndiaye, Anderson, Beto, Ekitiké -> Dúbravka, Van Hecke, Groß, Rice, Gordon, Haaland, Verbruggen, Calvert-Lewin, Guéhi	WC	Darlow, Raya, Andersen, Virgil, Enzo, Ndiaye, Wilson, Anderson, Beto, Ekitiké -> Verbruggen, J.Timber, Gomez, Groß, Gordon, Rice, Haaland, Dúbravka, Calvert-Lewin, Guéhi	WC	DIFF
33	Roll	BB	Roll	BB	=
34	Roll	FH	Roll	FH	=
35	Roll		Roll		=
36	Groß -> Sarr		Gordon, Gomez -> Wilson, Sarr		DIFF
37	Roll		Roll		=
38	Roll		Roll		=

Model Divergence

The projection divergences reveal v1’s systematic bias against players with zero elite ownership, most notably Haaland (-1.09), Welbeck (-0.88), and Gabriel (-0.76). All top 10 divergences show v2 higher than v1, with elite_eeo values at or near 0.00, indicating v1’s z-score methodology is penalizing these players despite strong baseline projections. Only Jarrod Bowen shows minimal elite ownership (0.04), suggesting the confidence-gated approach in v2 better preserves strong underlying fundamentals. This pattern explains why v2 achieves higher overall team projection while maintaining similar structural decisions.

Name	Pos	Team	v1_xP	v2_xP	delta
Erling Haaland	FWD	Man City	4.36	5.45	1.09
Danny Welbeck	FWD	Brighton	3.53	4.41	0.88
ID:791	FWD	West Ham	3.22	4.03	0.81
Viktor Gyökeres	FWD	Arsenal	3.18	3.97	0.79
Bukayo Saka	MID	Arsenal	4.70	5.49	0.79
ID:730	FWD	Sunderland	3.06	3.82	0.76
Gabriel dos Santos Magalhães	DEF	Arsenal	4.44	5.20	0.76
Jarrod Bowen	FWD	West Ham	4.21	4.93	0.72
Mads Hermansen	GK	West Ham	3.08	3.79	0.71
Daniel Ballard	DEF	Sunderland	3.82	4.53	0.71

Charts