Celtic Stats Corner: Rangers Preview and Regressing to the Mean
Going into the international break Celtic has taken maximum points from their first three matches, as well as qualifying for the Champions League Group Stages. The feel-good factor is back at Celtic Park, with fans and players alike talking about the positive atmosphere around the club since the arrival of Brendan Rodgers as manager. It seems to be the perfect time for Celtic’s first match up with Rangers.
In that same time, Rangers have dropped points twice in four matches. Draws against Hamilton and Kilmarnock, two clubs picked to be towards the bottom of the SPFL Premiership table, has led to frustration from the supporters of the club.
These struggles from Celtic’s Glasgow neighbors has led to Celtic supporters going into the September 10th matchup full of confidence. Yet, could the table be lying to us? Do the underlying stats suggest the two clubs are closer than we think?
First let’s take a look at Celtic’s stats. Celtic’s attack has been very impressive in the first three matches. They have an expected goal total of 5.88 and1.96 xG per game, which is the 2nd best in the league thus far, and an xG difference of 1 and 0.33 xG difference per game. Their total shot ratio, or how many shots a team is getting compared to their competition (i.e. Celtic have 10 shots and Aberdeen have 5 so Celtic’s TSR is 0.67 and Aberdeen’s is 0.33) is 0.61. Celtic’s TSR is also 2nd best in the league thus far.
Celtic also have a Total Shots on Target Ratio of 0.55 (same idea as TSR but with shots on target), which is 3rd best in the league.
All of these stats have led to 10 goals in 3 games for Celtic, the most in the Premiership. Celtic have taken 40 shots and 15 shots on target to get these 10 goals, which is a conversion rate of 25%. That seems pretty good right? It certainly is an impressive conversion rate, but unfortunately it is not sustainable.
Shots Against vs Shots For, @Ladbrokes Premiership 2016-17, @celticfc and @RangersFC allowing fewest shots so far pic.twitter.com/ROFYKYfz9D
— FitbaFancyStats???? (@226blog) August 29, 2016
Cumulative conversion rate vs cumulative shot total for 10 seasons of Scottish top div, note regression over time pic.twitter.com/ZseY37k2b7
— FitbaFancyStats???? (@226blog) September 2, 2016
Borrowing @226blog’s graph, we see Celtic’s conversion rate compared to the rest of the league. We also see a graph of conversion rate in the Scottish top flight for the past ten years. As you can see, both Celtic (green line) and the rest of the league historically regress in score rate. In other words, we can’t expect to see Celtic continue to convert shots into goals at such a high rate.
The good news is that while we cannot expect Celtic to continue to be so proficient in finishing, Celtic does have good underlying statistics to suggest their attack will still continue to thrive.
Having a good total shot ratio, total shots on target ratio and good xG numbers is promising for Celtic. These stats have shown to be better predictors of goals and points in the table than goals themselves.
The other end of the pitch might be more of a concern for Celtic. In their three matches, Celtic has conceded 4 goals or 1.33 goals per game, which is mid-table in the league. They have an expected goals conceded total of 4.88 or 1.63 per game, which is in the bottom half of the table thus far.
While Celtic have only conceded 8.33 shots per game and 4 shots on target per game, the high xG numbers mean that the shots Celtic’s opponents have been getting have been higher probability shots. That could be down to Celtic’s opponents thus far, all three finishing in the top half of the table last year.
Sun Tzu said, “if you know the enemy and know yourself, you need not fear the result of one hundred battles.” With that in mind, Rangers have dropped points in half of their matches so far, but also have the highest xG total and per game at 8.43 and 2.11 respectively. They have also earned the highest xG difference (3.29), highest TSR (0.67) and TSoR in the league (0.67).
In their four matches, they have scored 6 goals on 66 shots and 20 on target, for a conversion rate of 0.091.
More from The Celtic Bhoys
If Celtic’s conversion rate is not sustainable at 25 percent, then we can also expect Rangers conversion rate to go up a bit towards the mean. The good expected goal and shot numbers suggest Rangers have been a bit unlucky in their goal scoring so far. Could those numbers be bolstered by the fact three of their first four opponents were bottom six teams in the table last year?
Yes, that is very possible. But Rangers attacking performance statistics were numbers you would expect them to have against lower-half teams in everything but goals only.
In their first four matches, Rangers back line has faced intense criticism. That defense has allowed 4 goals, 5.14 xG (3rd best in the league) and 1.285 xG per game (2nd best in the league). They have the best shots conceded per game and shots on target conceded per game.
Again these numbers could be bolstered by the opponents they faced in their first few matches, but if these numbers stay where they are then we can expect Rangers defense to be among the better back lines in the league.
I enjoy laughing at dropped points by a team in light blue that plays at Ibrox as much as the next Celtic supporter, yet the early numbers suggest Celtic is not as far ahead of Rangers as most supporters believe.
However, Celtic have had three matches against some of the toughest opposition that the SPFL Premiership have to offer. They have taken maximum points against that opposition with solid attacking stats, though we can expect some regression in goal scoring.
There is no doubt Celtic will be the toughest opposition Rangers have faced this season so far. Rangers have had some misfortune against opponents from the bottom half of the table last year and have racked up impressive stats from those games.
I do believe that Celtic will be able to slow down that impressive attack from Rangers, but Brendan Rodgers and the Celtic support would be wise not to take their opponents on September 10th for granted.
This article originally appeared on 
