Another post from the archives; this time from February 2012.
The League Cup Final on Sunday afternoon got me thinking about Cardiff City, and when I think about Cardiff City, I always think of the last game of the 1923/24 season. That day must have been utterly wonderful, or utterly demoralising, depending on where you stand.
A little while ago I drew up a few tables and graphs about Town and Cardiff that season, and now seems as good a time as any to post them up. It’s a story well worth telling.
0.024 of a goal remains the tightest finish to a Football League Championship. Michael Thomas’ late second for Arsenal may have gone into recent folklore, but Len Davies’ penalty miss at Birmingham shaded the trophy even more narrowly into Huddersfield’s hands. It was, in truth, a trophy they probably wouldn’t have been expecting, being as many as six points behind Cardiff earlier in the season, and that in the days of two points for a win. After a draw on the 16th of February, however, Cardiff stalled. 43 points; Town had 37 at that point. When Cardiff next recorded a point, some 41 days later, they were two points behind Town, with the Terriers having won 4, drawn 2 and lost 1 of their seven games in the interim.
Here is a graph detailing the leadership between the top two at the end of the season; the important part mentioned above being the six horizontal arrows in dark blue. The 12th April (where the two coloured arrows meet) led to a topsy-turvy end to the season; first Town led, then Cardiff led then, as detailed, on the last day, Town pipped them by that brutally close margin.
The two teams were incredibly evenly matched, but there must have been something to suggest that Town would triumph? Well, as it happens, I’ve had a closer look, and there’s a few things that might point to a West Yorkshire victory overall.
The first divisor I looked at was the difference between home and away; both teams were far stronger at home than away, but Cardiff had less differential between their PPG for the two variables – 1.67 v. 1.05 for Town, 1.57 v. 1.14 for Cardiff); that difference was not represented in the games between the two – Town defeated Cardiff at Leeds Road and gained a goalless point at Ninian Park. The home / away split is the only one of the three that I looked at where Town were not the most consistent.
Having seen that breakdown of how points were gathered, I expected Town to have mounted a charge during the second half of the season, but the numbers don’t bear that out. 28 points over the first 21 games was bettered only by 1 in the second half of the season – 29; Cardiff’s 32 fell to 25 in the second half – which gave Town the impression of a charge, but – as I mentioned – Cardiff spent a long time treading water.
Another thing that people often level at teams who fail in the league is that they’re not good enough at beating either good teams or bad teams. Obviously, the top six contains the two teams in question, but Town were slightly better (0.2 points) better at winning against the top teams than Cardiff were, but 0.17 points worse at beating the bottom six – which balanced out, but again, Town were the more consistent.
So Cardiff started well, and did well against the lower placed teams but had a slower second half of the season, and were marginally less successful against the top six (it works out as 2 points difference) than Town did. That, to me, spells a team who weren’t the most consistent around. At the critical time, Town’s stubbornness and consistency won out, it seems, overturning Cardiff by the narrowest of margins. There are two more graphs here to look at to finish my points per game analysis, which is, first, a cumulative points per game tally, and then a last-6-games points per game tally.
I would expect Town’s to be relatively level (after the fluctuation of early games) but the trendline to rise slightly. Cardiff’s I expect to explode towards two, but then dip and dip. Meanwhile, the consistency of Town’s season (remember the graphs I drew comparing Town to Charlton a few weeks back?) I would expect to give a far more level graph and a lower standard deviation.
So, as it happened, it proved to be. I include trendlines in the 6-game form table; if only because its vaguely interesting to look where they intersect – game 20, if you can’t quite pan it down. The standard deviation (which is a relatively tenuous thing to use too often, but I think is quite revelatory in this instance alone) of Town’s 6-game form is 0.27, and for Cardiff is 0.40; which means what? That Town, over the course of the season, were more consistent than Cardiff by a measure of 0.13 deviations; so not much, but enough.
It seems to me, then, that the 1923/24 Football League Championship was decided between two teams – one who could beat any other on their day, and often did, and one who would get the same result against someone one day as they would the next. On this occasion, the ‘tortoise’ beat the ‘hare’.
Incidentally, Town closed the season with 2 wins, 5 draws and a loss; can you imagine a league champion finishing like that nowadays?