Will's Premier League Predictions

[willz121]

Round 5 - Actuals

Leicester_City 1 - 0 Southampton (H) = 40 points
Liverpool 0 - 0 Chelsea (D) = 0 points
Newcastle_United 1 - 1 Aston_Villa (D) = 40 points
Watford 1 - 2 Tottenham_Hotspur (A) = 0 points
Arsenal 0 - 0 Everton (D) = 0 points
Bournemouth 0 - 0 Burnley (D) = 0 points

Round 5 Total = 80 points

Decent haul in Round 5 with two perfect predictions, as Leicester secured a 1-0 home win over Southampton, and Newcastle and Villa shared the spoils in a 1-1 draw at St James Park. With a 19.1% prediction, the 1-0 home win for the Foxes came true, whist the 1-1 draw between Newcastle and Aston Villa was only 11.4%. Both teams here had a high prediction of scoring, which eventually came true with both teams finding the net in a 1-1 draw.

The stalemate at Anfield was impacted by the random request from myself for Chelsea, and the game ended in a 0-0 draw which only had a 8.3% chance of occurring. This scoreline was the fifth highest predicted, but the most surprising was Liverpool failing to score despite the 80.3% prediction.

Looking at the outputs from the model, the game at Vicarage Road was probably just in favour of a home win if the draw predicted was incorrect. The likelihood of a 1-0 or 2-1 home win were both higher, along with the 1-1 draw, than the actual predicted outcome but very pleased that the model did predict that both teams would score.

At the Emirates, the prediction that Everton wouldn't score came true, but surprisingly Arsenal also failed to find the net despite a 86.4% prediction that this would happen. What resulted was a lowly predicted 0-0 draw, especially when Arsenal were 72.1% favourites to win.

Finally, the biggest shock of the round was that the game between Bournemouth and Burnley finished goalless.

There should be goals here, kiss of death lol, with Over 2.5 goals reporting 68.6% suggesting it is more likely than not - enjoy the 0-0 draw chaps!

The likelihood of this result was just 2.9%, whilst both teams had a very high propensity to score - Bournemouth 75.0% and Burnley 88.4%! So of course based on those factors, the game had to end in a stalemate lol

Scores

Round 01 = 20 points
Round 02 = 10 points
Round 03 = 50 points
Round 04 = 10 points
Round 05 = 80 points (170 points total)

[willz121]

Round 6 - Predictions

Frustrating that I missed this post prior to the games being run, but the model predicted the following results:

Liverpool 2 - 1 Leicester_City
Southampton 1 - 1 Newcastle
Chelsea 1 - 2 Watford
Aston_Villa 0 - 1 Arsenal
Tottenham_Hotspur 1 - 1 Bournemouth
Everton 1 - 2 Burnley

The model was heavily in favour of Liverpool winning at home to Leicester in Round 6, with a 68.9% home win predicted, and Liverpool being 91.7% likely to score. Leicester were also likely to score at 64.9%, but the model still weighed in favour of a Liverpool win.

At St Mary's, the score draw was predicted with a 12.7% chance of happening. Both teams had goal predictions above 70% here, so the although the second highest predicted result was a 1-0 home win at 10.6%.

Battle of the basement clubs here, and the prediction is that Chelsea's poor start to the season would continue with the Hornets picking up a 1-2 win. Both teams had high propensity to score, Chelsea 81.8% and Watford 87.3%, which was reflected in the next highest predicted score was the 1-1 draw at 8.1%.

Arsenal were big favourites to win this tie, with a 50.4% win prediction. The predicted scorelines also showed that whilst Arsenal were likely to score, if Villa could grab a goal they would be in a great position to get at least a point.

The game at Tottenham saw the 1-1 draw a huge favourite at 10.9%. Both teams had identical predictions to score at 78.5%, and the highest predicted results were all BTTS in reflection of this. In the end, the 1-1 draw was the highest, although 2-1 either way was predicted second at 8.4%.

Finally to Goodison Park, where Burnley were once again predicted to secure three points on the road. With a win prediction of 56.4%, Burnley were huge favourites and despite their clean sheet in the previous game against Bournemouth, it was expected that Everton would breach the Claret's defence at least once. Burnley however had a very high propensity to score at 91.9%, which resulted in the final score prediction of 1-2 at 8.1%.

Again, apologies these weren't posted prior to the game, but I have written the predictions without any influence of the results so hopefully they provide interesting reading!

[willz121]

Round 6 - Actuals

Liverpool 2 - 1 Leicester_City (40 points)
Southampton 0 - 0 Newcastle_United (10 points)
Chelsea 1 - 1 Watford (0 points)
Aston_Villa 2 - 1 Arsenal (0 points)
Tottenham_Hotspur 3 - 1 Bournemouth (0 points)
Everton 0 - 0 Burnley (0 points)

Total = 50 points

At Anfield the 2-1 prediction was absolutely spot on, and Liverpool managed to make the most of their 91.7% chance of scoring. The away win was never likely on the cards, but a 1-1 draw was an interesting bet. However, Liverpool made it 2-0 before Leicester could get a consolation meaning they secured three points.

I predicted a draw between Southampton and Newcastle, and whilst it wasn't the 0-0 draw that occurred (7.8%), it's good to see that the model is starting to understand the promoted sides a bit better. Both teams had a good chance of scoring here, so it's interesting the game ended with neither side finding the net... there was a 92.2% chance that at least one goal would be scored!

The 1-1 draw at Stamford Bridge was rescued very late on by Dybala, and had Watford secured the 1-0 win, they would've defied the model with that scoreline only showing a 4.8% prediction. In the end, the 1-1 draw was a close second in terms of predictions (8.1%) against the 1-2 result (8.4%) that I predicted so it was pretty close!

At Villa Park, a home win was predicted at just 23.1%, and the 2-1 scoreline was 3.8%! A huge win here for Villa against an Arsenal side who are regularly predicted to do well and were massive favourites to win this tie.

If the Villa win at 3.8% was low, then the 3-1 win for Spurs was even lower, ranking at 2.8% propensity to occur. The 1-1 draw was the biggest prediction here, but Spurs turned on their form and managed to get a comprehensive 3-1 win. The 3-1 result would've been roughly 35/1 at the bookies, so a decent profit there if you could've backed it!

Finally, I once again provide the kiss of death on Burnley. At 8.1% to win 2-1 as predicted, the game finished goalless. Not only did both teams have a huge propensity to score (Everton 81.0% and Burnley 91.9%), but the chances of a 0-0 draw were just 1.5%! Second round in a row now for Burnley finishing in goalless draws!

Scores

Round 01 = 20 points
Round 02 = 10 points
Round 03 = 50 points
Round 04 = 10 points
Round 05 = 80 points
Round 06 = 50 points (220 points total)

[willz121]

Round 7 - Predictions

Leicester_City 1 - 0 Newcastle (20.1%)
Watford 0 - 2 Liverpool (12.2%)
Arsenal 2 - 1 Southampton (9.6%)
Bournemouth 1 - 0 Chelsea (15.8%)
Burnley 1 - 0 Aston_Villa (18.4%)
Everton 1 - 1 Tottenham_Hotspur (9.4%)

Once again, apologies that I missed the deadline for these, so I thought I'd keep it short and sweet and go straight to the results.

Round 7 - Actuals

Leicester_City 1 - 2 Newcastle_United
Watford 0 - 1 Liverpool (10 points)
Arsenal 1 - 0 Southampton (10 points)
Bournemouth 1 - 0 Chelsea (40 points)
Burnley 0 - 2 Aston_Villa
Everton 1 - 2 Tottenham_Hotspur

Newcastle totally destroyed the prediction here, with a 2-1 away win coming in at just 2.5%. This was the 10th most likely result, so for the Geordies to come away with the win here is a big surprise compared to historical data.

At Vicarage Road, I predicted a 2-0 win and Liverpool did deliver a win to nil, but only managed one goal. The 1-0 win wasn't far behind the predicted 2-0 win at 11.3% so very interesting to see the model going so close.

The historical data really expected Southampton to score here, but Arsenal kept a clean sheet and only needed the one goal to secure 3 points. Southampton had a 73% chance of scoring and BTTS was above 63% so a surprise here that the Saints didn't net, but Arsenal stood firm - the old 1-0 to the Arsenal!

Absolutely gutted this was predicted right as it meant we lost, and to be honest my caretaker reign at Chelsea has been absolutely terrible so far. But a perfect prediction as the Cherries secured a 1-0 win.

It must now be known as fact that I cannot predict Burnley results... once again I totally got this wrong as Aston Villa came away with a really good 2-0 win. This was only 3.9% likely to happen, but Villa come away with a very good result here!

Finally, Spurs went to Goodison Park and came away with a 2-1 win. I predicted a 1-1 draw which was a close run thing, but Spurs did enough to turn this around. Spurs were most likely to win of the two teams looking at the model, but the draw was the highest percentage likelihood.

A total of 60 points here so not too bad at all!

Scores

Round 01 = 20 points
Round 02 = 10 points
Round 03 = 50 points
Round 04 = 10 points
Round 05 = 80 points
Round 06 = 50 points
Round 07 = 60 points (280 points total)

[willz121]

Very far behind here...

Round 8 Predictions:

Watford 1 - 1 Leicester_City (11.7%)
Newcastle 0 - 1 Arsenal (13.9%)
Liverpool 1 - 1 Bournemouth (10.3%)
Southampton 1 - 1 Burnley (10.4%)
Chelsea 1 - 0 Everton (18.5%)
Aston_Villa 1 - 1 Tottenham_Hotspur (10.6%)

Weirdly lots of draws predicted in Round 8, and here are the actual results:

Watford 1 - 1 Leicester_City (40 points)
Newcastle_United 1 - 3 Arsenal (10 points)
Liverpool 2 - 0 Bournemouth
Southampton 0 - 3 Burnley
Chelsea 1 - 2 Everton
Aston_Villa 1 - 2 Tottenham_Hotspur

Not bad to collect 50 points from this round, with Watford and Leicester ringing true. The high percentage win predicted for Chelsea didn't come anywhere near true, and maybe that is down to their caretaker manager...

Scores

Round 01 = 20 points
Round 02 = 10 points
Round 03 = 50 points
Round 04 = 10 points
Round 05 = 80 points
Round 06 = 50 points
Round 07 = 60 points
Round 08 = 50 points (330 points total)

[willz121]

Round 9 Predictions:

Leicester_City 0 - 0 Arsenal (17.1%)
Bournemouth 1 - 1 Watford (12.0%)
Burnley 0 - 0 Newcastle (18.4%)
Everton 0 - 2 Liverpool (10.4%)
Tottenham_Hotspur 1 - 1 Southampton (10.6%)
Aston_Villa 1 - 0 Chelsea (17.1%)

A few draws again, and a couple of stalemates predicted at Leicester and Burnley. Actual results as follows:

Leicester_City 2 - 1 Arsenal
Bournemouth 1 - 0 Watford
Burnley 1 - 2 Newcastle_United
Everton 1 - 0 Liverpool
Tottenham_Hotspur 4 - 2 Southampton
Aston_Villa 0 - 1 Chelsea

Nothing to take away from this one, not even a correct outcome, and our first blank week!

Scores

Round 01 = 20 points
Round 02 = 10 points
Round 03 = 50 points
Round 04 = 10 points
Round 05 = 80 points
Round 06 = 50 points
Round 07 = 60 points
Round 08 = 50 points
Round 09 = 0 points (330 points total)

[willz121]

Round 10 Predictions:

Bournemouth 1 - 1 Leicester_City (11.3%)
Arsenal 1 - 1 Burnley (8.5%)
Watford 1 - 0 Everton (21.6%)
Newcastle 1 - 1 Tottenham_Hotspur (11.4%)
Liverpool 2 - 0 Aston_Villa (10.3%)
Southampton 0 - 0 Chelsea (17.7%)

Interestingly, the prediction for Watford to beat Everton 1-0 is 21.6% with the next most likely result being a 0-0 at 21.2%. Watford haven't won in their last five home games, scoring only twice. They've conceded an average of 2 goals per game, so a high prediction for a Watford clean sheet is very obscure.

Everton have only scored two goals away from home in the league all season, in their last away tie at Chelsea. Before that, they had two score less draws and a 2-0 defeat at Leicester.

Maybe the money is on a goalless draw (5/1), so watch out for goals here lol