Welcome to The Matchday Diget, here is everything you need for Gameweek 21 plus some bonus FA Cup content!

The biggest game of this gameweek is most likely to be Nottingham Forest vs Liverpool, a win could put Forest just 3 points behind Liverpool, however, the model sees this as unlikely with them just scoring 1.16 goals compared to Liverpool’s 2.45. This seems a bit high considering Forest’s defensive form as of late, however, they are playing the best team in the league.

Elsewhere the model gives Tottenham the lowest chance at a clean sheet in gameweek 21 as they play away to Arsenal. Currently, this sits at just 9.2%. The most likely clean sheet of the gameweek coming from Manchester United as they play a Southampton side who have scored just 5 goals away from home this season.

New for this week is the models ’ match win rate percentages! These are the results of my model simulating the game 10,000 times and recording how often each outcome occurs.

The 2 most sure results of the week according to the model would be Arsenal beating Tottenham at home and Newcastle continuing their run of form with a win vs Wolves.

In the biggest game of the week, Liverpool has just a 47.1% chance to beat Forest, with Forest themselves getting the win in 28.5% of the simulations.

Liverpool where as high as 92.3% after Arsenal drew 1-1 with Brighton but with them failing to beat a struggling Manchester United the chances have stayed pretty similar to last week. However, with Liverpool facing an in-form Forest, any more dropped points could add another massive swing in the chances.

A note here, Nottingham Forest for the first time has been seen in the model to win the league in just 60 of the 10,000 simulations (0.06%).

Southampton are now at a 99.9% chance of relegation according to my model. With them escaping the drop just 14 times out of 10,000…

Elsewhere Ipswich’s result vs Fulham meant that their survival hopes improved slightly with them now having a lower chance at the drop than Leicester City.

Wolves is one to look out for as they have a very difficult fixture run coming up which could see their chances increase dramatically.

Looking ahead

Here is proof of that incredibly tough run for Wolves, with them facing pretty much all of the best teams in the league this season over the next 6.

My ratings suggest Manchester United have the easiest next 6 currently, however, the Southampton game does bring down the ratings on average as you can see a Home game vs Southampton is rated at just 63.0, in comparison to Liverpool away as the hardest at 100.0.

Arsenal currently look best over the next 6 week period however expect this to look different next week as the model learns what Arsenal are like attacking-wise without Saka available as they are definitely not at full strength going forward.

However, defensively speaking they are still the best team in the league having just an xGA of 17.8 over 20 matches played which is just in front of Liverpool with an xGA of 17.7 over 19 matches.

Should the FA Cup be seeded from Round 3 onwards?

To answer this, I built a model which would simulate two different versions of the FA Cup. Both use all 64 teams present in Round 3 in 2025 with the same Elo ratings.

The seeds were all Premier League sides in 2023/24 and the top 12 teams from the Championship in 2023/24. Therefore Leicester City take seed 21 and Cardiff City take the final seeded spot at 32.

This means in theory, the top two seeded from last season Manchester City (1) and Arsenal (2) would if winning all their games only be able to meet in the final.

So what can we see?

A David vs Goliath match was defined as a game between teams which are at least 2 tiers apart. So a Premier League vs League One, Championship vs League Two and so on. With a Seeded FA Cup we would on average get 5.7 or 45% more of these games per FA Cup Third Round, meaning that lower league sides have a much larger chance to get that big away game vs a top side.

However, by Round 5, there is just an average of 0.4 David vs Goliath games compared to the unseeded format which has at least 1 on average.

Elsewhere we can see that it dramatically reduces the chances of a team in the third tier of English Football (League One) making it to Round 5 or further. In the seeded tournament on average, just 0.49 teams remain, however, in the Unseeded tournament, 1.46 are still here. Therefore it is almost 3 times as likely for a League One or lower team to still be in the tournament in the unseeded version.

As expected a seeded FA Cup would also protect the Premier League sides much more as of the 20 Premier League sides just 2.42 get eliminated in Round 3 compared to 5.12.

Finally, we can see that a Seeded FA Cup gives us 30 unique winners over 10,000 simulations, whereas the traditional unseeded FA Cup gives us 39. A 26.1% difference.

So while a Seeded FA Cup tournament would give lower teams a higher percentage chance at that big away ticket gate, it would significantly reduce their chances of advancing further in the tournament.