Transfer Uplift Explorer

Roman Bürki

GK · 8 apps · 720 min · 6.63 rating

↑ Upgrades

#	Player	Team	Apps	Rating	ΔWin%
1	Brad Stuver	Austin FC	44	7.04	+21.7%
2	Nicholas Hansen	Colorado Rapids	12	7.13	+15.4%
3	Dayne St. Clair	Inter Miami CF	41	7.06	+13.4%
4	Jun-Hong Kim	DC United	8	6.32	+11.1%
5	Hugo Lloris	Los Angeles FC	42	7.01	+9.4%
6	Aljaz Ivacic	New England Revolution	24	6.57	+9.4%
7	Michael Collodi	FC Dallas	19	7.01	+8.4%
8	Carlos Coronel	Red Bull New York	31	7.13	+8.3%

↓ Downgrades

#	Player	Team	Apps	Rating	ΔWin%
1	Jeffrey Gal	Chicago Fire FC	8	5.87	-10.0%
2	Ethan Horvath	Red Bull New York	8	6.65	-6.2%
3	Novak Micovic	LA Galaxy	24	6.54	-4.9%
4	John Pulskamp	Sporting Kansas City	42	6.39	-4.3%
5	Sean Johnson	DC United	39	7.2	-3.8%
6	Jonathan Bond	Houston Dynamo FC	33	6.59	-3.6%
7	Earl Edwards Jr.	San Jose Earthquakes	7	5.7	-3.5%
8	Evan Bush	Columbus Crew	6	6.75	-2.9%

52 candidates scanned against 7 opponents

Timo Baumgartl

CB · 8 apps · 720 min · 6.62 rating

↑ Upgrades

#	Player	Team	Apps	Rating	Goals	ΔWin%
1	Christopher Cupps	Chicago Fire FC	6	6.45	0	+1.9%
2	Morris Duggan	Minnesota United	32	6.73	1	+1.6%
3	Kalani Rienzi	Seattle Sounders FC	28	7.04	4	+1.6%
4	Keegan Rosenberry	Colorado Rapids	23	6.68	3	+1.5%
5	Tim Parker	Red Bull New York	15	6.98	1	+1.3%
6	Julian Gressel	Minnesota United	26	6.71	1	+1.3%
7	Jefferson Díaz	Minnesota United	37	6.9	1	+1.2%
8	Finn Surman	Portland Timbers	43	6.88	0	+1.2%

↓ Downgrades

#	Player	Team	Apps	Rating	Goals	ΔWin%
1	Alvas Elvis Powell	FC Cincinnati	35	6.53	0	-1.3%
2	Dany Rosero	Sporting Kansas City	5	6.64	0	-1.2%
3	Kamal Miller	Portland Timbers	33	6.56	2	-1.2%
4	Joseph Rosales	Austin FC	36	6.78	1	-1.2%
5	Julio Cascante	Austin FC	15	6.43	0	-1.1%
6	Joel Waterman	Chicago Fire FC	33	6.85	1	-1.1%
7	Malte Amundsen	Columbus Crew	35	7.07	1	-1.1%
8	Tomás Avilés	CF Montreal	31	6.52	2	-1.0%

201 candidates scanned against 7 opponents

Daniel Edelman

MF · 8 apps · 692 min · 7.08 rating

↑ Upgrades

#	Player	Team	Apps	Rating	Goals	ΔWin%
1	Milan Iloski	Philadelphia Union	34	7.04	14	+9.6%
2	Evander Ferreira	FC Cincinnati	43	7.73	19	+9.0%
3	Jeppe Tverskov	San Diego FC	46	7.78	2	+5.7%
4	Sebastian Berhalter	Vancouver Whitecaps	42	7.69	7	+5.6%
5	Aleksey Miranchuk	Atlanta United	40	7.1	10	+4.6%
6	Cole Bassett	Portland Timbers	32	6.97	3	+4.4%
7	Rodrigo De Paul	Inter Miami CF	23	7.35	2	+4.0%
8	Pep Biel	Charlotte FC	34	7.3	14	+3.9%

↓ Downgrades

#	Player	Team	Apps	Rating	Goals	ΔWin%
1	Erik Duenas	Houston Dynamo FC	8	6.54	0	-3.7%
2	Allan Oyirwoth	New England Revolution	5	6.49	0	-3.2%
3	Show	FC Dallas	6	6.29	0	-3.1%
4	Stephen Afrifa	Sporting Kansas City	16	6.41	0	-3.0%
5	David Ruiz	Inter Miami CF	6	6.48	0	-2.7%
6	Brandon Servania	DC United	37	6.62	0	-2.6%
7	Memo Rodriguez	Sporting Kansas City	19	6.33	0	-2.5%
8	Wil Trapp	Minnesota United	45	6.84	2	-2.5%

233 candidates scanned against 7 opponents

Christopher Durkin

MF · 8 apps · 688 min · 7.3 rating

↑ Upgrades

#	Player	Team	Apps	Rating	Goals	ΔWin%
1	Milan Iloski	Philadelphia Union	34	7.04	14	+11.4%
2	Evander Ferreira	FC Cincinnati	43	7.73	19	+11.2%
3	Jeppe Tverskov	San Diego FC	46	7.78	2	+5.7%
4	Sebastian Berhalter	Vancouver Whitecaps	42	7.69	7	+4.3%
5	Joaquín Pereyra	Minnesota United	45	7.35	7	+4.2%
6	Rodrigo De Paul	Inter Miami CF	23	7.35	2	+3.9%
7	Mateusz Klich	Atlanta United	16	6.55	0	+3.7%
8	Sergio Busquets	Inter Miami CF	39	7.4	0	+3.6%

↓ Downgrades

#	Player	Team	Apps	Rating	Goals	ΔWin%
1	Show	FC Dallas	6	6.29	0	-3.6%
2	Erik Duenas	Houston Dynamo FC	8	6.54	0	-3.3%
3	Allan Oyirwoth	New England Revolution	5	6.49	0	-3.1%
4	Maxime Dominguez	Toronto FC	24	6.68	0	-2.8%
5	Matias Rojas	Portland Timbers	8	6.77	1	-2.8%
6	Stephen Afrifa	Sporting Kansas City	16	6.41	0	-2.8%
7	Ian Harkes	San Jose Earthquakes	37	6.74	3	-2.8%
8	Pedro Soma	San Diego FC	12	6.26	0	-2.7%

233 candidates scanned against 7 opponents

Simon Becher

ST · 8 apps · 629 min · 6.87 rating

↑ Upgrades

#	Player	Team	Apps	Rating	Goals	ΔWin%
1	Heung-Min Son	Los Angeles FC	20	8.0	12	+33.5%
2	Lionel Messi	Inter Miami CF	41	8.48	42	+27.1%
3	Hugo Cuypers	Chicago Fire FC	41	7.49	25	+21.0%
4	Denis Bouanga	Los Angeles FC	42	7.8	30	+17.8%
5	Sergi Solans	Real Salt Lake	7	7.68	5	+17.6%
6	Nicolás Fernández	New York City FC	24	7.57	11	+13.5%
7	Guilherme Augusto	Houston Dynamo FC	7	8.35	5	+12.2%
8	Petar Musa	FC Dallas	40	7.57	28	+10.5%

↓ Downgrades

#	Player	Team	Apps	Rating	Goals	ΔWin%
1	Ali Ahmed	Vancouver Whitecaps	27	7.26	1	-5.1%
2	Ahmed Qasem	Nashville SC	37	6.4	2	-4.9%
3	Pedrinho	FC Dallas	25	6.65	2	-4.9%
4	Magomed-Shapi Suleymanov	Sporting Kansas City	40	6.69	3	-4.8%
5	Caden Clark	DC United	36	6.42	0	-4.7%
6	Tucker Lepley	LA Galaxy	16	6.64	0	-4.7%
7	Wiki Carmona	CF Montreal	5	7.33	3	-4.7%
8	Wikelman Carmona	Red Bull New York	36	6.55	0	-4.7%

216 candidates scanned against 7 opponents

Dante Polvara

CB · 7 apps · 623 min · 7.04 rating

↑ Upgrades

#	Player	Team	Apps	Rating	Goals	ΔWin%
1	Michael Boxall	Minnesota United	37	6.83	1	+2.8%
2	Mamadou Fofana	New England Revolution	36	6.9	1	+2.6%
3	Tim Parker	Red Bull New York	15	6.98	1	+2.5%
4	Kevin Long	Toronto FC	24	6.8	0	+2.4%
5	Omar Gonzalez	Chicago Fire FC	12	6.84	0	+2.2%
6	Oleksandr Svatok	Austin FC	39	6.59	1	+2.1%
7	Christopher Cupps	Chicago Fire FC	6	6.45	0	+2.1%
8	Gilberto Flores	FC Cincinnati	23	6.87	0	+2.0%

↓ Downgrades

#	Player	Team	Apps	Rating	Goals	ΔWin%
1	Joseph Rosales	Austin FC	36	6.78	1	-2.7%
2	Max Arfsten	Columbus Crew	40	7.53	7	-2.5%
3	Julio Cascante	Austin FC	15	6.43	0	-1.2%
4	Alvas Elvis Powell	FC Cincinnati	35	6.53	0	-1.1%
5	Julián Aude	LA Galaxy	35	6.71	0	-1.0%
6	Bill Poni Tuiloma	Charlotte FC	8	6.78	1	-1.0%
7	Kamal Miller	Portland Timbers	33	6.56	2	-0.9%
8	Carlos Harvey	Minnesota United	32	6.92	2	-0.9%

201 candidates scanned against 7 opponents

Jaziel Orozco

CB · 7 apps · 615 min · 6.6 rating

↑ Upgrades

#	Player	Team	Apps	Rating	Goals	ΔWin%
1	Kalani Rienzi	Seattle Sounders FC	28	7.04	4	+2.5%
2	Julian Gressel	Minnesota United	26	6.71	1	+2.4%
3	Christopher Cupps	Chicago Fire FC	6	6.45	0	+2.2%
4	Omar Gonzalez	Chicago Fire FC	12	6.84	0	+2.2%
5	Morris Duggan	Minnesota United	32	6.73	1	+2.2%
6	Tim Ream	Charlotte FC	38	6.77	0	+2.0%
7	Devin Padelford	Minnesota United	22	6.59	0	+2.0%
8	Matt Miazga	FC Cincinnati	25	6.83	0	+1.9%

↓ Downgrades

#	Player	Team	Apps	Rating	Goals	ΔWin%
1	Joseph Rosales	Austin FC	36	6.78	1	-1.4%
2	Kamal Miller	Portland Timbers	33	6.56	2	-1.1%
3	Alvas Elvis Powell	FC Cincinnati	35	6.53	0	-1.0%
4	Malte Amundsen	Columbus Crew	35	7.07	1	-0.7%
5	George Campbell	CF Montreal	16	7.03	0	-0.6%
6	Osvald Gabriel Søe	San Diego FC	5	5.72	0	-0.5%
7	Alan Montes	Sporting Kansas City	8	6.23	0	-0.4%
8	Joel Waterman	Chicago Fire FC	33	6.85	1	-0.4%

201 candidates scanned against 7 opponents

Rafael Santos

FB · 6 apps · 458 min · 6.86 rating

↑ Upgrades

#	Player	Team	Apps	Rating	Goals	ΔWin%
1	Jordi Alba	Inter Miami CF	36	7.41	6	+5.2%
2	Andrés Herrera	Columbus Crew	8	7.04	1	+1.0%
3	Richie Laryea	Toronto FC	25	6.93	3	+0.6%
4	Nouhou Tolo	Seattle Sounders FC	37	7.04	1	+0.5%
5	Griffin Dorsey	Orlando City	39	6.59	2	+0.5%
6	DeAndre Yedlin	Real Salt Lake	37	6.92	1	+0.1%
7	Souleyman Doumbia	Charlotte FC	11	6.77	0	+0.0%
8	Jonathan Dean	Chicago Fire FC	37	6.7	1	-0.0%

↓ Downgrades

#	Player	Team	Apps	Rating	Goals	ΔWin%
1	Guilherme Biro	Austin FC	40	6.93	4	-2.4%
2	Nathan Harriel	Philadelphia Union	36	7.26	2	-2.1%
3	Benji Kikanovic	San Jose Earthquakes	6	7.22	0	-2.0%
4	Jimer Fory	Portland Timbers	36	6.96	0	-1.9%
5	Luca Bombino	San Diego FC	38	7.08	2	-1.9%
6	Tate Johnson	Vancouver Whitecaps	38	6.89	1	-1.9%
7	Daniel Lovitz	Nashville SC	41	6.97	2	-1.9%
8	Mathías Laborda	Vancouver Whitecaps	39	7.25	8	-1.8%

74 candidates scanned against 7 opponents

Conrad Wallem

FB · 5 apps · 447 min · 6.8 rating

↑ Upgrades

#	Player	Team	Apps	Rating	Goals	ΔWin%
1	Jordi Alba	Inter Miami CF	36	7.41	6	+2.7%
2	Kai Wagner	Philadelphia Union	35	7.86	2	+0.8%
3	Jonathan Dean	Chicago Fire FC	37	6.7	1	+0.8%
4	Kyle Duncan	Minnesota United	26	6.97	1	+0.8%
5	Brooks Lennon	Atlanta United	25	6.6	1	+0.8%
6	Nouhou Tolo	Seattle Sounders FC	37	7.04	1	+0.7%
7	Ilay Feingold	New England Revolution	36	7.0	4	+0.7%
8	Raheem Edwards	Toronto FC	31	6.91	0	+0.7%

↓ Downgrades

#	Player	Team	Apps	Rating	Goals	ΔWin%
1	Andy Nájar	Nashville SC	43	7.22	2	-2.3%
2	Tate Johnson	Vancouver Whitecaps	38	6.89	1	-2.0%
3	Benji Kikanovic	San Jose Earthquakes	6	7.22	0	-1.9%
4	Luca Bombino	San Diego FC	38	7.08	2	-1.7%
5	Richie Laryea	Toronto FC	25	6.93	3	-1.7%
6	Guilherme Biro	Austin FC	40	6.93	4	-1.4%
7	Édier Ocampo	Vancouver Whitecaps	43	6.96	3	-1.3%
8	Jasper Löffelsend	San Diego FC	12	6.79	0	-1.3%

74 candidates scanned against 7 opponents

Marcel Hartel

AM · 4 apps · 360 min · 7.47 rating

↑ Upgrades

#	Player	Team	Apps	Rating	Goals	ΔWin%
1	Luciano Acosta	FC Dallas	21	7.31	5	+6.7%
2	Carles Gil	New England Revolution	41	7.84	11	+4.3%
3	Diego Luna	Real Salt Lake	32	7.29	11	+3.4%
4	Thomas Müller	Vancouver Whitecaps	20	7.72	12	+3.4%
5	Amine Bassi	Houston Dynamo FC	29	6.84	2	+3.2%
6	Djordje Mihailovic	Colorado Rapids	39	7.48	14	+2.3%
7	Facundo Torres	Austin FC	8	7.36	1	+2.0%
8	Diogo Goncalves	Real Salt Lake	30	7.0	4	+1.7%

↓ Downgrades

#	Player	Team	Apps	Rating	Goals	ΔWin%
1	Griffin Yow	New England Revolution	6	6.47	1	-0.4%
2	Cheikh Sabaly	Vancouver Whitecaps	6	6.89	1	-0.3%
3	Antony	Portland Timbers	39	7.14	8	-0.1%
4	Jesús Ferreira	Seattle Sounders FC	42	7.0	4	+0.7%
5	Zavier Gozo	Real Salt Lake	33	7.17	6	+0.8%
6	Cameron Harper	Red Bull New York	12	6.98	3	+0.9%
7	Emil Forsberg	Red Bull New York	41	7.11	12	+1.1%
8	Dagur Thórhallsson	CF Montreal	5	6.45	0	+1.2%

17 candidates scanned against 7 opponents

Sergio Córdova

ST · 6 apps · 270 min · 6.75 rating

↑ Upgrades

#	Player	Team	Apps	Rating	Goals	ΔWin%
1	Heung-Min Son	Los Angeles FC	20	8.0	12	+38.7%
2	Lionel Messi	Inter Miami CF	41	8.48	42	+35.1%
3	Hugo Cuypers	Chicago Fire FC	41	7.49	25	+34.4%
4	Joseph Paintsil	LA Galaxy	29	7.28	11	+28.9%
5	Petar Musa	FC Dallas	40	7.57	28	+28.8%
6	Denis Bouanga	Los Angeles FC	42	7.8	30	+27.5%
7	Sergi Solans	Real Salt Lake	7	7.68	5	+27.4%
8	Kelvin Yeboah	Minnesota United	42	6.79	14	+26.8%

↓ Downgrades

#	Player	Team	Apps	Rating	Goals	ΔWin%
1	Ali Ahmed	Vancouver Whitecaps	27	7.26	1	-3.8%
2	Caden Clark	DC United	36	6.42	0	-3.5%
3	Ahmed Qasem	Nashville SC	37	6.4	2	-3.5%
4	Omari Glasgow	Chicago Fire FC	14	6.05	0	-3.5%
5	Ryan Kent	Seattle Sounders FC	16	6.7	0	-3.4%
6	Wiki Carmona	CF Montreal	5	7.33	3	-3.3%
7	Wikelman Carmona	Red Bull New York	36	6.55	0	-3.3%
8	Magomed-Shapi Suleymanov	Sporting Kansas City	40	6.69	3	-3.3%

216 candidates scanned against 7 opponents

Concept Visualized

Pipeline

📊

Match Stats

65,000+ player-match rows
31 raw features

Goals, rating, xG, tackles, passes…

→

🧠

Autoencoder

31 → 32 → 31

Compresses stats into
32-dim player embeddings

→

⚽

Position Buckets

GK · CB · FB · MF · AM · ST

Embeddings summed per
position → 192-dim team

→

🎯

XGBoost

77% accuracy
calibrated

Predicts P(Win), P(Draw),
P(Loss) from 384 features

Uplift Calculation

Baseline

Team lineup → team emb → P(Win)

vs

Swap

Replace starter with new player
at same position → new P(Win)

=

ΔWin%

Win probability change
ranked across all candidates

How This Works

Methodology

Step 1: Learn Player Representations

An autoencoder (a type of neural network) is trained on ~65,000 player-match stat rows from 2022-2026. It learns to compress 31 raw features (goals, assists, rating, xG, tackles, progressive passes, etc.) into a 32-dimensional vector that captures how a player performs in a single match. Players with similar stat profiles end up close together in this embedding space — even if they've never played against each other.

Step 2: Build Team Embeddings by Position

For each match, players are grouped into position buckets: GK, CB, FB, MF, AM, ST. Their embeddings are summed within each bucket, then concatenated into a 192-dimensional team embedding. This means the model treats a team as the sum of its positional roles — not just a bag of players. A team's defensive strength is captured independently from its attacking threat.

Step 3: Predict Match Outcomes

An XGBoost classifier is trained on concatenated [home team emb | away team emb] = 384 features to predict home win, draw, or away win. Trained on 2022-2025 data, tested on 2026. The predicted probabilities are calibrated with isotonic regression so they reflect true win rates rather than overconfident raw scores.

Step 4: Measure Transfer Uplift

For each starter in the lineup, every position-compatible MLS player is tested as a replacement. Their embedding swaps into the team's positional bucket, and the model predicts the new win probability against each opponent. The difference (ΔWin%) is the marginal value of that transfer. Results are compared against a null baseline of 30 random players at the same position to flag statistically significant picks.

What ΔWin% Means

A +5% uplift means "this player makes the team 5 percentage points more likely to win on average." In a 34-game season, that translates to roughly 1.7 additional wins. The null baseline tells you whether that signal is meaningful — if the p95 threshold for that position is +2%, then a +5% uplift is above what you'd expect from any random player at that position.

Limitations

This model captures statistical profile only — not chemistry, tactics, coaching fit, leadership, or locker-room dynamics. It can't model how a player adapts to a new system or whether they'll click with new teammates. The positional buckets are broad (e.g., all fullbacks are grouped together). Treat these numbers as a signal, not a verdict.

Top Signings

Roman Bürki

Timo Baumgartl

Daniel Edelman

Christopher Durkin

Simon Becher

Dante Polvara

Jaziel Orozco

Rafael Santos

Conrad Wallem

Marcel Hartel

Sergio Córdova

Concept Visualized

How This Works