Points Per Cost II: What’s Value for Money at 2018/19 prices?

After last week’s initial Points Per Cost article, many people asked to see this year’s prices compared to last year’s performance.

Whilst we initially weren’t too sure about doing this, with us initially thinking the idea of forcing an equivalence between this year’s prices and last year’s prices would be wrong. Our thinking here was that the underlying assumption that good performance = price rises might not be true. For example, Willian was outscored by Cesar Azpilicueta by 40 points last season, but saw a 0.5 price rise whilst his Spanish team-mate saw a 0.5 fall – there seemed no logic in that.

However, thinking slightly outside the “researcher” box, we did begin to wonder what this might look like. Another way of looking at the data that would seem odd might be that surely those popping by comparing this season’s prices to last season’s performances would indeed be “value for money” (i.e. indicating a price drop).

We did this by dividing their cost now by their performance last season. We also compared metrics with and without appearance points, with the latter our preferred measure (as we did in the prior article).

As well as fishing for value, we also get to investigate whether good performance does indeed correlate with price rise decisions.

Let’s see what we find out per position…

This year’s data v last year’s performance – Goalkeepers

(Top: Last year’s points v last year’s prices)

Lukas Fabianski is far and away the best value player in terms of providing “value” for your money through delivering points above appearance. He’s been in the top three for saves for the past three seasons among keepers. This tallies with our article on goalkeeper choice, which suggested the optimal yield for a goalkeeper relative to price is to have a cheap keeper paired with a premium defender (e.g. Fabianski paired with Chris Smalling).

It’s interesting to see how, based on last year’s performances, David De Gea (DDG) is still third for points per cost despite the 0.5m price hike. This shows the value for money that a bit of an outlier season (25% more clean sheets; 39% more saves than an average of the prior three campaigns) can give you. This will surely normalise to around fellow expensive keeper Ederson between 15-16 points per cost.

Another name along with Fabianski who pops is Brighton’s Mat Ryan, who kept 10 clean sheets and made an average of 3.26 saves per game last season, meaning you tended to get an added 1 point from owning the Australian per game. That’s pretty decent too, and one who might be flying under the radar due to an adverse couple of fixtures in Gameweeks 2 and 3 against Manchester United and Liverpool.

This year’s data v last year’s performance – Defenders

(Top: Last year’s points v last year’s prices)

Now here’s where things begin to get interesting.

Chelsea pair Cesar Azpilicueta and Marcos Alonso seeing a price of 6.5, compared to a hike for Nicolas Otamendi and Ben Davies, see them both sit best value for money amongst defenders. If either (with Azpilicueta surely nailed on) look a dead cert to start next season it may be worth paying them heed and parting with the extra cash to get them in your side. “Dave” topped the bonus points chart last season by some distance (in effect, you started every game with 2 points and were likely to get bonus if Chelsea kept a clean sheet) whereas Alonso topped attempts for a defender last season. Either could be a devastating weapon in your squad; best to wait until the Community Shield with them though.

Interestingly, we see Andrew Robertson fall off if we roll for this year’s value to just below Kieran Trippier, who actually played 38 minutes less than the Scotsman did. This may make those who, like us, spot value in Robertson have cause for concern. However, this is a point where the stats require context. This is because the full back only really got in gear at the tail end of last season – four of his five assists came between Gameweeks 28-38 when he was playing the vast majority of the games. This means he’s more of an “emerging pick” based on end-of-season form rather than one based on a whole-season analysis – which is fine. Patrick van Aanholt may also fall into this category.

Once again there are no defenders cheaper than 5.5 in this top 10 – showing that the “value” lies in the top end for the position.

This year’s data v last year’s performance – Midfielders

(Top: Last year’s points v last year’s prices)

No surprises who sits top, as King Mo Salah’s insane season shows through him still retaining top value despite the 4.0m price hike to 13.0m! Similarly, Raheem Sterling sits second after his 18 goals and 17 assists last season earnt him 229 points and a 3.0 price hike – given a reminder of what he achieved last year, I’m now less surprised that he earnt the price shunt but there was still initial shock to see him cling on to his second spot.

In fact, in terms of the midfield, five of the top six spots are Manchester City players. This highlights what a strong season they had last season, and, perhaps, the conundrum that many FPL managers face with their fixtures looking amazing between Gameweeks 2-7. David Silva‘s minutes being managed may mean some look at Leroy Sane, who was only beaten by a wafer thin margin by the Spaniard, as the most reasonable City option. New signing Riyad Mahrez outscores Kevin de Bruyne and keeps his spot through costing 1.0 less, but we’ll need to see if he can carry on his consistently decent output at a bigger club.

We see a couple of players come into the top 10 with appearance removed, too. Luka Milivojevic and Xherdan Shaqiri fall from the chart, perhaps indicating they’re less good value than their price tags suggest. This leads to giving Christian Eriksen and Aaron Ramsey spots in the top 10, though it’s worth noting the Dane only marginally outscores Talisman Theory hero Pascal Gross in terms of points per cost.

This year’s data v last year’s performance – Forwards

(Top: Last year’s points v last year’s prices)

A bit of a surprise here, as Jamie Vardy parties his way up to the top of the points per cost metric as his 9.0m pricing contrasts with the 1.0 price hike Roberto Firmino had to take him up to 9.5m. 20 goals last season saw the England man net 183 points – I was surprised to be reminded that Vardy two points more than Firmino did last season. Interestingly, Firmino got the 1.0 hike whereas Vardy only got 0.5, perhaps showing how team status has an impact on pricing as, objectively, they both (as they started last season on 8.5) should be priced near enough the same in terms of price-to-performance calculations.

Vardy becomes interesting around Gameweeks 5-15, when his Leicester have a purple patch of fixtures which sees them only face Arsenal of last year’s top six. I’d expect the rodent-featured striker to hit nine or ten goals in that period, meaning he may well be my “go-to” striker if I Wildcard around Gameweek 6.

This data for forwards, as Nick mentioned in his previous article, and also his “Death of the Third Striker” piece, is very underwhelming. Only Vardy and Firmino make it into the top 50 in terms of value for money using this metric.

However, after the inclusion at 7.0 of Marko Arnautovic and Wilfried Zaha in this category, as well as Cenk Tosun starting his first full season, it might be that this points per cost recovers after a fallow year last season. As it stands, though, two strikers only is in vogue with a 4.5 striker taking third bench in both of our sides. Alexander Sorloth is taking some interest as a first bench choice, too.

Here’s the top 20 for Points Per Cost v this season’s prices – excluding appearance.

Conclusion

The first thing to say is that this proved that pricing significantly correlates with performance to a huge degree across the players currently in the game. Reassuringly, this shows on a quantitative level that prices are mostly set by objective means, rather than human ones. Our underlying assumption that prices are led by performance is shown to be almost totally proven.

… almost totally.

Analysing the price in this way did indeed throw up some interesting individuals that pop in the data we’ve seen. Think the Chelsea defenders and Vardy v Firmino. This could be explained by specific weights being put on by team strength, perhaps. We’ve seen in the past that team performance can impact pricing – naturally, after Leicester’s title winning 2015/16 season, their players started on average 1.5m over priced.

At times, team performance may be overriding player pricing rules based on their output as individuals – which helps us spot value. The early season tends to be a exercise in finding value players, adding grist to many a social media conversation and inspiration for analyses such as this one.

Identifying those players and investing early is what we all want to do. Looking at players without appearance points giving a nice way to find those with a historic record of “value add” in terms of what they give you. The data shows that cheap ‘keepers, premium defenders, Man City midfielders and, erm, Jamie Vardy look the best points per cost.

In the end, value is in the eye of the beholder to some extent. The example with Andy Robertson shows that although there may only be value in specific periods rather than season long. Observing the trends at the end of the season (and showing awareness of the limitations of a season-long analyses) are key to helping interpret value beyond the metrics – remember that this is near enough objective value measurement, but you apply your interpretation of value when choosing your team.