We are reaching the decisive moments of the season and, from now on, every small decision can have massive implications in our Overall Rankings and Mini League standings.
In this article, I will go through the analytical approach I like to use when I am debating if I should take a hit or not, by looking into my latest decision as an example.
I hope it can help you whenever you face a close call like the one I faced last weekend.
Blank Gameweek 29 went really well for my team. Despite only fielding nine players, I managed to get 62 points and one of the highest Gameweek ranks of the whole season.
Joel Veltman (£4.4m), Jesse Lingard (£6.2m) and Patrick Bamford (£6.7m) delivered double-digit hauls and my captain Harry Kane (£11.6m) converted a penalty against Aston Villa. That propelled me back into the top 6k after a couple of disappointing Gameweeks in a row.
Getting Kane and Lingard for Dominic Calvert-Lewin (£7.6m) and Mohamed Salah (£12.4m) with my two free transfers were fairly easy and standard decisions. Kane was a must-have given his massive ownership in the top 10k and I really wanted Lingard, who seemed to be enjoying life under David Moyes’ command, with four goals and two assists in his first six matches for West Ham.
Having decided what to do with my transfers, I started wondering if taking a hit would be worth it, in order to add one more player to my already depleted squad. Son Heung-Min (£9.4m) and Bamford were yellow-flagged, meaning there was a chance I would only field seven players, which was a bit less than I was hoping for when I started planning for the Blank Gameweek several weeks ago.
After looking at all possible options, I came to the conclusion the only hit that could possibly make some sense, would involve getting Lewis Dunk (£5.0m) in for João Cancelo (£6.1m).
Brighton and Hove Albion was, by far, the most likely team to get a clean sheet so I would not mind a defensive double-up, by adding Dunk to Veltman (who I bought in the previous Gameweek).
At the same time, I knew I was going to use my second Wildcard ahead of Gameweek 31 so I did not really need Cancelo anymore since he was blanking in Gameweek 29 and I already had a decent defence for Gameweek 30 anyway.
But how to know for sure if getting Dunk was worth throwing away four of my hard earned points? Personally, I like to remove all bias and gut feelings from my decisions whenever possible by using an analytical approach to figure out if the player I am considering buying will get me over or less than four points on average.
In order to decide if I was buying Dunk or not, I made a list of all the ways Dunk could win or lose points:
- Playing time
- Clean Sheet
- Attacking returns
- Bonus points
- Two goals conceded
- Yellow cards
Then, I tried to figure out how many points I could expect from Dunk in each of these actions individually. Let’s go through them.
PLAYING TIME
This one was easy. Being ever-present at the heart of the Seagulls defence, Dunk was guaranteed to play over 60 minutes, so I would be shocked if he did not get me two points just for playing.
90 minutes = 2 points
CLEAN SHEET
For determining how likely it is for a team to keep a clean sheet I like to use clean sheet odds, which I think is the most accurate metric available, even if it might be a bit subjective.
According to G-Whizz’s weekly hot topic on clean sheet odds, Brighton had a 48% chance of keeping a clean sheet against Newcastle United. If we multiply it by four (number of points a defender earns for getting a clean sheet), we see that Dunk would get 1.92 clean sheet points on average:
0.48*4 =1.92 points
ATTACKING RETURNS
To analyse Dunk’s potential in terms of attacking returns, I decided to use data from the last three seasons. I always prefer to use long-term numbers whenever I think the old data is just as reliable as the new one (which is usually the case for centre-backs).
So here are Dunk’s attacking numbers since 2018/19:
– GOALS PER 90 – 0.09
– ASSISTS PER 90 – 0.04
Having assembled this data, the only thing left to do is to multiply these numbers by the number of points that defenders get for goals (6 points) and assists (3 points):
(0.09*6) + (0.04*3) = 0.54 + 0.12 = 0.66 points
BONUS POINTS
Dunk got a total of 35 bonus points in the last three seasons, having played a total 8,631 minutes, which means he gets 0.36 bonus points per 90 minutes played:
(35 / 8631) * 90 minutes = 0.36 points
TWO GOALS CONCEDED
According to the sports betting industry, Brighton had roughly a 20% chance of conceding two or more goals against Newcastle and it seemed like a fair number to me, so I went with that:
0.20 * (-1) = -0.20 points
YELLOW CARD
Dunk has seen a yellow card 18 times in his last 8,631 minutes, which means he loses 0.19 points per match on average due to his fouls:
( 18 / 8.631 * 90 ) * (-1) = -0.19 points
CONCLUSION
Finally, by adding all the average points Dunk would get for each of those actions, we see that the Brighton captain was worth a total of 4.55 points on average for his Gameweek 29 match against Newcastle.
This means taking a hit to get him in, would give me a net gain of 0.55 points on average according to the above calculations and so, I decided to trust my method and go for it.
Even though it seems like a small gain, we face several close decisions like this over the course of a season and they certainly add up to our overall points. Making the best decision on average whenever we face close decisions can be the deciding factor between finishing inside the top 10k or not, between winning a mini-league or finishing second just a couple points behind. For this reason, I like to make sure I am taking every small thing into account by using an analytical approach instead of just trusting my gut or following the template.
In the end, Brighton did keep a clean sheet and Dunk got six points, which means he over-performed what my calculations expected by 1.45 points.
But even if he blanked, I would be happy with my decision of taking a hit, knowing that if I keep making transfers with a positive expected value in the future, it will work out more times than not.
3 years, 1 month agoEnjoy Fabio's writing as always.