Relegation watch revisited (2): It was the draws wot done it

John McCormick:
John McCormick
number crunching

The season before last we won only 7 games and we stayed up. We even finished above Aston Villa, who won 10, as well as QPR and Hull City, whose 8 wins each could not stave off relegation. The other relegated team, Burnley, had 7 wins, the same as us.

The difference between us and Burnley was that we achieved 17 draws, and lost only (only??) 14 games. They could manage only 12 draws, and their five fewer points meant they finished second bottom, three places below us.

And thinking about that got me started on the notion of win-loss ratios, which became the tool I used to track clubs in last season’s relegation watch.

In particular, our 7 wins and 14 losses got me wondering if a club which lost exactly twice as many games as it won would always finish with 38 points. And that got me wondering if any club with 38 points must have lost exactly twice as many games as it won, so I worked it out mathematically and found they must, at which point Malcolm pointed out that one win and two losses is equivalent to three draws so of course they must.

But by then I was running with my idea. A win-loss ratio of 1:2 (or 0.5) would result in 38 points and that’s usually enough for safety so I made it my starting point. Then I dropped the ratio slightly to reflect the points gap that often separates the bottom three from the rest (ignoring 2010-11, when two teams went down with 39 points) and made my prediction:

Any club with a win-loss ratio of 0.48 or better would stay up. Any with 0.45 or worse would go down.

Between those figures, in the zone of uncertainty, clubs might stay up or go down but would not control their own destiny by the end of the season. To survive they would have to depend on other clubs losing.

The clubs I chose to follow were newly promoted Norwich, Watford and Bournemouth,  along with Leicester, Villa and Newcastle, the bottom clubs in the premiership. And, of course, Sunderland was down there with them. Over one thousand fans voted on which three of them would go down but I was tracking all seven from the start of the season, not that I needed to track Leicester for long.

You can see the final win-loss ratios in the chart below on the left:

win_loss rations for the relegation candidates Leicester’s magnificent ratio (23 wins, 3 losses) calls for a long axis and makes it difficult to separate the clubs in the huddle around the 0.5 mark, so I’ve taken Leicester (and Villa) out and expanded the axis around this key region:

Win-loss ratios, selected clubs
Win-loss ratios, selected clubs

You can see that Watford, Bournemouth and Sunderland were comfortable by the end, as befits clubs in control of their destinies, and even we achieved safety with a game to spare.

NUFC finished in the zone of uncertainty. They could have stayed up, especially after Rafa started a resurgence, but only if we did worse than them in the final run-in, and we didn’t.

Norwich also lost control of their destiny. The crunch games were the two six-pointers. Winning against Newcastle kept Norwich in the fight but our win at Carrow Road left us one point behind them with a game in hand. From then they were reliant on us dropping more points than they did. As it happened, they lost three of their last four games, and finished with a win-loss ratio well below the 0.45 relegation boundary.

Comfortable by the end? Maybe so, but we were well out of the comfort zone for virtually the whole competition. We bounced along in the bottom three until Sam’s arrival brought an improvement, but winter brought five losses  in succession, which left us seven points adrift of safety. With the turn of the year, on the back of those losses, our win-loss ratio had fallen to 0.23. Three wins from nineteen games is relegation form, no doubt about it, especially when your mini-league competitors have all staged revivals.

The chart below shows the direness of our position.

I’d said any club consistently below 0.45 would go down. The only teams consistently below 0.45 were us and Villa:

Win-loss ratios, games 6-38

Watford were out of sight, with Bournemouth not far behind them. Norwich had taken themselves above the 0.48 line and even the Mags had got above it once. I felt then, more than at any other time in the last four years, and not because of my maths, that we wouldn’t dodge the drop, even though we had the best team we’d assembled since 2012. That’s how bad things were.

And then we picked up, Norwich went on a slide and the Mags faltered. We won a couple of games but just as importantly we also drew games we would have lost earlier in the season. Sam and Sixer bemoaned the dropping of points when we couldn’t keep a clean sheet. They could be justified in doing so but bear this in mind – in the first half of the season we drew three games. In the second half we drew nine 

And the draws made a difference.  We also improved our ability to win, doubling our tally, and we can’t underestimate the importance of those wins, but the draws kept us going. (And how crucial was that one at Mr Ashley’s SPORTSecond DIRECTivision Arena?).

Newcastle’s second half of the season matched their first, thanks mainly to Rafa, who reversed a terminal decline, and they added five wins to the four they gained before New Year. Norwich were much poorer after December. They did win four games, near enough matching their five in the first half of the season, but they lost thirteen, as opposed to only nine of the first nineteen games.

As a result, by the end of the season, all three teams had won 9 games, but our win-loss ratio was better because of our draws.

Unlike Newcastle and Norwich, losses weren’t dragging us down.

Our ratio had risen above 0.5, whereas Norwich were hovering around 0.4 while Newcastle, who had been well below Norwich at one point, had surged to within touching distance of 0.48 only to come up just that little bit short. Ah, well…

…too little, too late for Newcastle; a stunning end to the season for us.

But not one I’d like to repeat



12 thoughts on “Relegation watch revisited (2): It was the draws wot done it”

  1. I think we need an article on Squad Analysis. There are those who simply must go. Those who could go, but also stay, and areas we need to strengthen.
    Personally, I think we need 1) Decent Cover at the back, as in Kaboul & Kone we trust. A new Right Full Back, as the one we have is not wonderful, but Sam may help him. Yedlin back please. Midfield only needs creative players. And attack we need someone young to challenge Jermaine. Hardest position to fill. We need a hungry 20 / 21 years old wanting to prove himself that no one risks on and we take. Not Connor Wickham style of course. But we need that sub who will be the next Defoe.

    • How much are we paying Rodwell? Whatever it is it’s money down the drain in my opinion, canny lad but not up to it, get what we can for him and say adios.

      • I have no problem with Rodwell. What has he done wrong: when he has had a real opportunity?

        Get him in for the pre-season, get him wound up, let him go.

        And if he doesn’t do it then, the let him go.

      • What has he done right John? He disappears from games, often only visible when he gets booked unnecessarily, and is always injured. OK let’s see if Sam can get him up to speed with a full pre-season and if he still doesn’t cut it get rid. I’d love him to prove me wrong but I’ve seen little to justify his large fee, probably massive wages, and the many chances he’s been given to prove himself.

      • He has at times been pretty decent, though I agree these performances have not been as frequent as we might wish.

        He came in for a lot of stick when he missed a couple of decent chances in front of goal but in his defence he had got himself into a position where they were decent chances.

        Cattermole, Kirchhoff and M’Vila looked a solid partnership in midfield but the one weakness in that line up was they rarely got themselves into goalscoring positions and were generally reduced to long range efforts which rarely troubled the opposition.

        I still think he will be a decent member of the squad if he can stay fit. When you think how much Yedlin, van Aanholt , Kaboul and Defoe’s contributions improved following Allardyce’s appointment there’s no reason to think the coaching staff can’t improve Rodwell’s input too.

  2. Thanks John. Brilliant analysis.

    I think the key thing, from our point of view, is that Mr Short brought in Sam early enough to effect change. I think that had Dick Advocaat had lasted six more games, it would have been too late.

    Newcastle on the other hand, gave McClaren, six games too many. I think, if Rafa had had those extra games, that they may have stayed up [ their last game against Spurs demonstrated the progress he had made ]

    Villa and Norwich stuck with what they had, understandably in Norwich’s case, but blindly, I feel, in Villa’s.

    Next season will present new challenges, but I have a strong feeling that we will not be scrapping to stay up next April. Hope not anyway. Can’t handle too many more .

  3. I kept looking at our goal difference. We were better than our rivals for a long time and I did feel that this would even itself out points-wise by the end of the season – i.e. we’d get unexpected points and stay up as our GD suggested we deserved to.

    • I’d need to check but I think that we had a really poor goal difference at Christmas. Villa were worse but if I remember correctly Norwich were better and the Mags no worse

Comments are closed.

Next Post