6.515 – Another Point About Substitutes – Post 42

It struck me, while discussing the number of substitutes in League One yesterday, that I might be missing something quite obvious; its about those time-wasting substitutions at the end of games, and how they might skew the figures. I know, for example, that Paul Tisdale played about a minute for Exeter at the end of their season, and his one appearance (which was about 90 seconds long) counted in the figures. So I needed to try something that would balance it out a bit more; I was drawn, as often in these circumstances, to percentages.

Basically, what percentage of a teams’ playing time was taken up by substitutes? There’d be a maximum of somewhere around 27% (3 players coming on for 90 minutes a time) but I didn’t expect we’d trouble double figures, to be honest. What I didn’t expect to find is that Huddersfield Town topped the table for time played by substitutes, with 6.515% of the total league game time being taken up by replacements – this figure being in no small part due to Jordan Rhodes’ seeming inability to fulfil his 90 minutes whenever he played and the willingness of Lee Clark to remove Benik Afobe from the fray.

My thinking was, then, that the more successful teams would use their substitutes later, just to waste time, and the worse teams would use them sooner, so as to try to change things. I was, as yesterday, misguided.

Look at the percentage line (in red) dipping massively from almost six to just over 4.5, despite the wild inconsistencies (Sheffield Wednesday, with their huge squad, showing up there) from individual team to team. I took into account, for this, the total minutes played (11 x 90 for each game, less the minutes lost for sendings off) and balanced that into percentages with the total time played by the substitutes. Those of you who read yesterday’s post won’t be surprised to see MK Dons featuring so highly; the top three being Town 6.515, Sheffield W 6.432, and MK Dons on 6.430. It’s a fair bit, that, really; its about 65 minutes of game time every match being played by a substitute. Because this table is a bit more compact, I’ve put it in at the bottom. Enjoy reading it, and well done to Oldham and Dagenham, who had the most faith in their starting XIs last season, or the least faith in their substitutes, depending on how you sway. The three columns, if you’re interested, are total minutes (including playoffs), total substitute minutes, and % of Total Minutes that were played by Substitutes.

2.63 – Substitute For Another Point – Post 41

There are eleven players on a football team, aren’t there? Yes and no. Football, even at League One level, is increasingly a squad game. We saw it in action at Huddersfield last season, when the home formation and the away formation were both entirely different. This leads to the knock-on effect of squad players, and by virtue of that, substitutes being more valuable.

This summer sees the reduction of substitutes from seven back down to five, something I’m in favour of. I don’t think that teams need to have squads of 18 for league games, and it’ll balance things more in the favour of the smaller teams; no disrespect to them, but it seemed there was always someone being shown up last season for not having a squad big enough to field the full complement.

That said, this post is about use of substitutes, and my hypothesis would be that the better teams would use fewer substitutes because the games would already be going their way, and they wouldn’t need to change things as often during the games – though there is the gap-

narrowing ‘time-wasting’ substitution towards the end of the game.

The reality however, was the exact opposite. The better the team, the more players were replaced. There is a point here that the better teams aren’t just better XIs; they’re better throughout the squads, so a Southampton substitute would be more likely to effect a game positively than a Bristol Rovers substitute. This led to the top six averaging 2.68 substitutions per game (top mark to MK Dons, who made 2.83 – that works out as MK Dons only leaving eight substitutions unused throughout the season – a staggering figure). The bottom four (points-gaining teams; Plymouth don’t feature by virtue of having 10 deducted) averaged only 2.21 substitutes per game.

If I was losing games, I’d like to think I’d try to change it rather than keeping faith with something that isn’t working, which is what seems to have happened last season; Swindon, as you can see from the graph, made an attempt to do so, but they can’t change a pretty defined line of best fit (red line) taking the amount of used substitutes downwards in conjunction with the league table.

All that said, there is another side of this coin is that there is a completely different looking line of best fit (green line) in the table of how many players each team used last season – again, I thought the better teams would use more players, as there’d be more availability to bring in younger players, and test them in a successful outfit. Not so. The desperation for points must spread into the pre-game bath. The line is very definitely slanted upwards in terms of players used by the worse-performing teams (look at the dip to Rochdale, by the way – their 24, and Exeter’s 25, players used is an incredible feat given the positions they both finished in). Squad sizes, then, are getting bigger, in League One, and its not always to the benefit of the teams who have them – better to spend money on quality than quantity, says the numbers, and its hard to disagree.

The new development regarding numbers of subs available will give these graphs an extra dimension next season – it might well be worth shadowing in the lines of best fit to see if we’re averaging more or less now there’s a smaller pool to play with. I think – not that MK Dons can do much about it – we’ll see even more substitutes this coming season, and even more players like Novak and Tehoue who make more than 20 appearances off the bench.