RED, the University of Reading’s Forecasting Model lay dormant over the summer. But it woke up just in time for the start of the English Football League season this weekend. The Model’s nature has been upgraded too, and has now become Bivariate-Poisson-AutoregressiveX, with the new bit being the “AutoregressiveX” (More on this in a later blog post or research paper).
This upgrade appears to have been seamless, but RED is coming up with some fairly controversial predictions now, though we hope it will settle down over the next few weeks:
The Model’s forecasts for R1 (2-5 Aug) of the English Championship are in the table below:
| Scoreline | Result Prob. | ||||
| Most Likely | Prob. | Home | Away | ||
| Luton | Middlesbrough | 2-1 | 0.10 | 0.61 | 0.15 |
| Barnsley | Fulham | 2-1 | 0.10 | 0.65 | 0.13 |
| Blackburn | Charlton | 1-1 | 0.10 | 0.41 | 0.31 |
| Brentford | Birmingham | 2-1 | 0.10 | 0.68 | 0.11 |
| Millwall | Preston | 1-1 | 0.11 | 0.43 | 0.29 |
| Nottm Forest | West Brom | 1-1 | 0.10 | 0.38 | 0.34 |
| Reading | Sheff Wed | 1-1 | 0.10 | 0.36 | 0.35 |
| Stoke | QPR | 2-1 | 0.10 | 0.64 | 0.13 |
| Swansea | Hull | 2-1 | 0.10 | 0.59 | 0.16 |
| Wigan | Cardiff | 1-1 | 0.10 | 0.48 | 0.25 |
| Bristol C | Leeds | 2-1 | 0.10 | 0.58 | 0.17 |
| Huddersfield | Derby | 1-1 | 0.10 | 0.26 | 0.47 |
- The point forecast (most likely) Scoreline, with the Probability of that happening
- The Probability density forecast of a win by the Home or Away team, with one minus those two numbers giving the probability of a draw