← Back to Blog

Sorare Next5 Case Study

Putting our money where our model is

We’ve just released our Next5 Projections for both Sorare and FPL. So I wanted to run a little experiment by actually building a team based on its findings and see how it does over the next 5 weeks. Over the next season, really.

The only issue is: I don’t have a lot of money. So this team will be build on a BUDGET (where possible). Ok, it doesn’t actually end up being a budget team, I got carried away. Let’s call it a VALUE team, because there are some good picks in there. . I’ll be running this experiment in Limited, but would love to do another one later in the season in Rare.

And remember, our model uses incredibly deep data analysis to run hundreds of thousands of simulations based on every team and player to find out exactly which fixtures and matchups have an edge. It’s far more complex than traditional fixture difficulty charts or even DFS “defence vs position” metrics. You can find out more about how our model works here.

But the model is only providing the projections. A human is picking the team: and that human is me. So results may vary. That’s also the beauty of Fantasy Matchup Edge… We provide the data for everyone, but there’s loads of different ways you can run with it. Different budgets, different playstyles, and hundreds of different ways to run with the data.

So, onto the team…

I already have Erling Haaland and Emiliano Martinez. I bought one of those special offer packs at the start of the casino because I don’t have enough self control. We’ll price them as what they’re available for NOW for this experiment, and I won’t include any other players from my gallery for this experiment other than the ones we buy for it.

Haaland

Haaland is already an incredible fit for this team. He’d be someone I’d seriously be considering bringing in even if I didn’t already own him. He’s ranked 4th overall on Sorare for the next 5 Gameweeks, although City play Utd first, and he’s only 9th for that GW. Still, his average 90th percentile (P90) ceiling score over 100,000 simulations per GW is 97.91. And that’s exactly what we’re looking for (more on that in a bit). Priced at £90.

Martinez

The GK spot was always going to be tricky, and expensive. So it’s nice that Martinez was included in the pack I bought. At the time, he was benched for Villa and looking for an exit, so becoming first choice at Chelsea landed well (especially for the first GW). He will have gone up in price a lot since then, but to be fair, we’ll be pricing him at today’s value since I’m supposed to be building this team RIGHT NOW. He’s actually the 7th best GK on our Next5 board, with an average expected Sorare score of 49.90. Priced at £64.

But these guys aren’t “budget” enough

Correct. We’ve spent a lot so far. And, having looked closely at our projections, there’s one other spot where it might actually be sensible to spend big again. But that’s not what this experiment is supposed to be about. We need to find some VALUE for this team, that could be competitive compared to “better-known” (and more expensive) players.

So I did a little research:

I quickly took our Top Next5 Projections to try and narrow down the field into some possibilities. At this point, FME’s projections are “if starts”. We model everyone based on what we know about them. This will be getting improved later to starters/probables only on our main public-facing sheets, but for now, it includes injured players and other people who are highly unlikely to play. So I need to narrow down the field, as well as looking at what players are available for on Sorare RIGHT NOW.

I also ranked our Top 100 picks by Avg P90. That’s their average 90th percentile ceiling score across the next 5 Gameweeks. The reason? I want guys who have the potential to BANG rather than just high average, low variance players.

Here’s what I came up with:

John McGinn

Now, the clearest value there for actual starts, for me, is John McGinn. He’s only £7. The only issue is that they might rest him after Champions League midweek. But he’s ranked 27th on the list, average expected score over Next5 at 60.57 and Avg P90 88.09. No brainer, I’m locking him in.

John McGinn – Priced at £7.

Reece James

Reece James is another good pick from that list. The only issue is that he’s changed position slightly, the model clearly knows more about how he USED to play, and while it’s still learning and will improve on things, current projections are still a bit unclear. Still, at that price, I think it’s value. Locking him in.

Reece James – Priced at £14.

So far we’ve got a GK, DF, MID, and FOR. That means we’re pretty free to buy who we want. Now, the real choice is whether to buy Bruno or not. He is NOT good value on a points per cost level, but he is also by far and away the clear outstanding Sorare player on the board. His Average P90 Ceiling is 100. And he is expected to get 2.64 80+ scores over the next 5 Gameweeks, which is clear of any other candidate, by far. But he’s expensive. Can we really call this a budget team if we include him? Is it really a budget team anyway? We’ve already got Martinez and Haaland.

Bruno Fernandes

Bruno Fernandes – Priced at £96.

I’ve decided we want a sub. And in case anyone gets injured soon. He needs to be cheap, so Marmoush stands out. But so does Kiernan Dewsbury-Hall. KDH has slightly higher average means, but Marmoush has a slightly higher chance of BIG scores, so I’m going for Marmoush.

Omar Marmoush

Omar Marmoush – Priced at £18.

So that’s the team:

• Martinez
• James
• McGinn
• Bruno
• Haaland
• Marmoush

Total cost: £289 (eek!)

Let’s see how they do. We’ll be back with a look after the GW, and looking forward to the following GW. Remember, we won’t have any collection bonuses or anything else, so this is a bit of a gamble. It was always a gamble. But that’s football.