Updated Projected League Table: Rangers move up & keep an eye on Killie
written by: @TheGersReport
This is part of an ongoing series in which I will regularly share a projection model that makes data based final predictions of the Scottish Premiership table. It is rooted in projecting goal differentials for the season. The model banks points already earned by each club & applies Expected Goal differentials to project out the rest of the season. For more, check out this introductory post.
It’s safe to say I didn’t do the best job of rolling out my new projection model last week. The intent was to introduce the origins of the model & highlight how teams could use it to pinpoint areas of improvement.
That’s it. I like writing about how data should drive internal analysis at clubs.
The actual words from the blog post were meant to be disclaimers of putting too much stock the small sample size of four weeks…but let’s just say the Rangers Twitterverse didn’t take too kindly to the initial projections.
Look, it was a rough couple of days on Twitter…which means I didn’t do a great job of really highlighting the fact that this model will see a great deal of change in the first month or two.
However, as more matches are played:
Teams will have banked more (or less) actual points & that will be reflected in the projections
The Expected Goals Differentials will become more solidified
This means the projections will build accuracy as each week of the season passes. So…yes the projections will feel more accurate in March or November then they do in September…
But, you know what….I’m still going to publish the projections….now after five matches because:
- They can help teams highlight areas of improvement (like I discussed last week as it pertains to Rangers)
- They can help track the form of a team (I see you Celtic)
- They can create talking points for blogposts like this one; i.e. why is Kilmarnock projected to be in the top three??
When it comes to form, you can see that Rangers & Livingston are trending in the right direction. Last week, Rangers’ projected points were 45.20 (hence the vitriol online) & now it’s up to 57.46. Livingston's win, & continued strong play, saw their projections go from 46.42 to 53.62.
I don’t think we’ll see such drastic differences week-to-week as more matches are played, but teams should always be looking to evaluate their form - not only in results but also in the actual performance on the pitch. This model combines those factors.
It’s also interesting to take note of the team currently projected to be in third place. In “real life” Kilmarnock sit seventh in the league table with seven points…so why does the model have them sitting third ahead of the likes of Hibs & Rangers?
Last week I highlighted the fact that since the 2009-10 season there was a 94% correlation between goal differential & points earned. I ran similar studies only looking at goals for & goals against & while the correlation wasn’t as spot-on, there was still real relationship between scoring goals (or preventing goals) & earning points.
Below, you'll find Kilmarnock’s projected points when you isolate goals for & allowed, along with Expected Goals & Expected Goals against so far this season.
What’s driving Killie’s projected success is their defensive play. Both their actual & Expected Goals allowed numbers are of a club poised to be one of the top teams in the league this season.
Some numbers to highlight how good Killie’s defence has been this season:
- Are one of two teams to have only allowed one goal from open play this season (Celtic is the other)
- Has the lowest Expected Goals against total in the league (3.25)
- Have won a league high 68% of their tackles this season
They still have some work to do to make the attack more effective (their projected points on Expected Goals is worrisome).
Killie’s current Conversion Rate of 10.9% is basically equal to their Expected Conversion Rate of 10.6 & they are still sitting on a pretty uninspiring 1.2 goals per match.
However, the team has the right pieces in place to be a dangerous team in attack (Kris Boyd, Jordan Jones, the full backs, Eamonn Brophy….and look it’s Greg Stewart).
And remember…it’s only five matches. It’s a small sample size… these are not FINAL predictions for the league table. The projected table mentioned above is only spitting out predictions from a model that will get much more accurate as more games are played.
This post was written with the aid of StrataData, which is property of Stratagem Technologies. StrataData powers the StrataBet Sports Trading Platform, in addition to StrataBet Premium Recommendations. Additional stats are courtesy of InStat Football
I came under some criticism for creating “clickbait” with the headline that Rangers may struggle to finish in the top three. A couple of points: I really don’t think my writing has been clickbait material over the past three years…I mean Scottish fitba analytics doesn’t necessarily go viral. Also, I do believe it will be hard for Rangers to finish in the top three. Part of that concern comes from the fixture demands of playing in the Europa League (a trade-off the club should be very happy to live with) & a bigger factor is the fact that Hearts, Hibs & even Kilmarnock will be very competitive foes this season in the fight for second (& third place). Shit, as each day passes since the departure of Stuart Armstrong…the more vulnerable Celtic’s claim of dominance looks. The point is Rangers are among a group of five talented teams led by very good managers all competing for those top spots in the league table. So, yeah…it’s going to be a struggle. #clickbait
The next four league fixtures are really intriguing for Rangers. Their next four matches are against St. Johnstone (who may actually be…good), Livingston (are you not believers yet?), Hearts (ohhh, that’s going to be fun), & Hamilton (actually worse then Dundee). How Rangers look in the projected table after all that is going to be very telling.
This was written under the influence of the afterglow of a 2-2 draw against Villarreal, the VSS, & Slint.