Projecting the 2018-2019 SPFL Table with B.U.R.L.E.Y., Our SPFL Premiership Model
Written by Matt Rhein/@thebackpassrule
For the third season running, we have fired up the gigantic computer that houses our SPFL model B.U.R.L.E.Y. Whenever I write this post each year, even I have to check what I decided B.U.R.L.E.Y. would be short for. As a reminder for us all, B.U.R.L.E.Y. stands for “footBall Using Reliable anaLytics, Even You!” Yes, it's a stretch of an acronym just as it was three years ago! I was inspired by Craig Burley's diatribe against expected goals and was going to force his name into the acronym! Despite forcing an acronym, I have generally been happy with B.U.R.L.E.Y.'s results these past few seasons handicapping what the SPFL table will look like at season's end.
If you are interested in how B.U.R.L.E.Y. comes up with his calculations, you can skip to the last few paragraphs. I get into the nitty gritty math, so if you like poisson distributions, monte carlo models and the such, you can find that nerdy stuff at the end. If you just want to see where B.U.R.L.E.Y. puts your club so you can complain how "stats are pish, we're finishing top 6", step right up!
With all that procedural stuff out of the way, above is how B.U.R.L.E.Y. sees the SPFL Premiership table come May 2019. Most would agree with B.U.R.L.E.Y.'s champion pick of Celtic. The top 4 gets a bit more controversial with B.U.R.L.E.Y., as he sees Rangers finishing second 11 points behind Celtic. He then has Aberdeen third, 10 points off of Rangers. In fourth, he has Hibs, 5 points off of Aberdeen. B.U.R.L.E.Y. sees Kilmarnock and Motherwell rounding out the top 6.
In the bottom six, B.U.R.L.E.Y. cares not if you find his results controversial, putting Hearts ninth after finishing sixth last season. The rest of the bottom six is not too radical, though not for the first time, B.U.R.L.E.Y. is picking Hamilton to go down. Each time, the Accies have proven the robot and many pundits wrong. Can they do it again? He also has the two newly promoted clubs St. Mirren and Livingston in tenth and eleventh.
Hearts fans will surely take issue with the ninth place projection. The Tynecastle club has altered their squad significantly in the off-season after their sixth place finish last season, which has lead to optimism from Hearts fans. B.U.R.L.E.Y. does not factor new signings so he does not factor all the additions to Hearts added this season in his projections, but the above tweet makes a good point. The underlying metrics for Hearts last season were not anything spectacular. They largely performed above those metrics, so B.U.R.L.E.Y.'s ninth place prediction is largely banking on a regression from the club. Could the new signings help negate that possible regression? Sure, but more often that not in these instances it does not happen.
Will your club outperform B.U.R.L.E.Y.'s projections? They could! B.U.R.L.E.Y. certainly was not expecting Killie to finish fifth last season. Yet, please save your "you're biased against my club" tweets. B.U.R.L.E.Y. is a bunch of poisson distributions and monte carlo simulations. He has no feelings about your club. Don't @ him.
To come up with the B.U.R.L.E.Y. SPFL Projected Table, I borrowed Mark Taylor, from the Power of Goals Blog, simulations method with a few tweaks. Using the expected goal data from the SPFL from last season, I took the xG average for the league, the average xG both home and away for every club, and the xG for and against for every club to come up with a calculation for xG for every match up between every team in the SPFL Premiership.
With these expected goal figures, I use Poisson distribution to come up with the probability of every scoreline for every match up in the league. If we sum the winning score lines for each team in each of these match-ups, we can determine the probabilities of who B.U.R.L.E.Y. thinks will win each match.
Once we have the win probabilities for every match up, we can then run simulations for every game each club has remaining in Excel, as Mark details how to in the link above. We run each club’s remaining season 1000 times. We take the average point total B.U.R.L.E.Y. suggests will happen and boom, that’s how many points B.U.R.L.E.Y. thinks your team will end up with.
I have also made a few tweaks to he B.U.R.L.E.Y. formulas the last few seasons. Before last season, teams at home in the SPFL typically outperformed their opponents by 10% in underlying metrics, so I gave the home club of every match up a bit of a boost in B.U.R.L.E.Y. At the conclusion of last season, that average jumped to 18%, so clubs playing at home have been given an even bigger boost in B.U.R.L.E.Y.
In the past few seasons, the SPFL club returning to the Premiership after earning promotion had resources and a fan base much larger than a typical promoted club. Therefore, we could not treat those clubs as your typical promoted club. This is not the case in St. Mirren and Livingston this season though. Therefore, since we do not have Championship data to base our model's predictions on for those clubs, we based their data on clubs of similar stature the last three years in the Premiership. It is not a perfect fit, but for lack of better options it is what we are going with.
This article was written with the aid of StrataData, which is property ofStratagem Technologies. StrataData powers the StrataBet Sports Trading Platform, in addition to StrataBet Premium Recommendations.