FanTeam

1 week ago

Fanteam's Monster [GW36]: Goalkeeper Value Analysis

Football is not a high-scoring type of sport, meaning we don't have a privilege of multiple projections for each player to make our fantasy lifes much easier. Due to FanTeam's football scoring, goalkeepers is one position that is more prone to more or less accurate point projections compared to other positions. In this article, I go over the methodology to calculate saves numbers and potential value for each goalkeeper of the slate. If you find the math too heavy on you, just scroll down to the very bottom of the page for an easy-to-read chart that will help you pick the right goalkeeper based on your team composition.

 

Calculating Goalkeeper Projections

First of all, we have clean sheet odds, chances to concede 1.5+ goals and chances to win from bookies at our disposal. Also, there is a lot of useful information about saves, which is the most important aspect of Fanteam GK scoring (each save is worth 0.5 points).

I predict the number of saves for every goalkeeper in the slate based each team's game style and possible possession %. 

We can't just simply take the average number of saves for each goalkeeper and use it for every slate because there are different types of matches. Every team can play differently depending on the location, the opponent, form, etc., which can influence how many saves a goalkeeper is likely to make. 

This model for predicted saves is based on possession. If we correctly predict how a team is likely to play their next game, we can use the save data from previous matches that were similar in terms of possession. 

For my analysis I single out 3 possession categories:

• team has >55% possession;
• team has 45%-55% possession;
• team has <45% possession.

I predict possession percentages based on previous performances of the current season.

Now that we have a prediction of how each team is likely to perform against their current opponent, I also have to mention that it is football we are talking about.

Xhaka can get a red card in the 7th minute (lol, I hope it won't be an exact projection, because the Gooners will hate me for that) or Wolves can score an early goal and games will play out differently. We can't see into the future, but this method gives us enough grounds to try and predict the number of saves. 

Pred SoT MIN — a minimum number of predicted shots on target based on stats through this season 
Pred SoT MAX — a maximum number of predicted shots on target based on stats through this season 
Pred Goals — predicted goals based on bookie odds
PSoT-PG MIN  Pred SoT MIN - Pred Goals
PSoT-PG MAX — Pred SoT MAX - Pred Goals

I have analyzed the stats based on projected possession and taken the number of shots on target for each team just from the matches falling within the same possession category.

I've also considered the current form of each team and where the game is going to be played. In the next part of our analysis, I will look at opponent stats. 

Each shot on target can become a goal, so we should exclude predicted goals from this sheet, leaving us with just potential saves. 

If we're talking about predicted goals, it is based on bookies odds. When we subtract predicted goals from predicted shots on target, we get potential saves. Here we determined an attacking potential and we will go back to these numbers later, but now we have to analyze the defensive potential of each team.

Predicted Save MIN — a minimum number of predicted saves based on stats through this season 
Predicted Save MAX — a maximum number of predicted saves based on stats through this season 

I did the same thing with saves here. Selected matches with similar possession percentages played by a given team this season and looked at how many saves each goalkeeper had made in those.

Now that we have a range of predicted saves based on the team's defensive potential, let's do a little math and merge the two. 

PSoT-PG MIN  Pred SoT MIN - Pred Goals
PSoT-PG MAX — Pred SoT MAX - Pred Goals
Save MIN — a minimum number of predicted saves based on stats through this season
Save MAX — a maximum number of predicted saves based on stats through this season 
APS MIN — the average number between PSoT-PG MIN & Save MIN
APS MAX — the average number between PSoT-PG MAX & Save MAX

Let's look at the first row of the sheet above so that I can explain how it works. 

[0.00 - 2.84] is the predicted range of shots on target for BUR (AVL's opponent).
[1.00 - 4.00] is the predicted range of saves for Martinez.
[0.50 - 3.42] is the average between the two previous ranges and my final projection, where we consider the attacking potential of the 1st team and the defensive potential of the 2nd team.

If we put it simply, my calculations show that Martinez is likely 1 to 3 saves against Burnley. 

Now let's take other aspects of GK scoring on Fanteam into account. 

CS (%) — chances to keep a clean sheet
FP (CS) — predicted Fantasy Points for a clean sheet (+4)
APS MIN — the average number between PSoT-PG MIN & Save MIN
APS MAX — the average number between PSoT-PG MAX & Save MAX
FP (APS MIN) — predicted Fantasy Points for Average Predicted Saves (MIN, +0.5 for each)
FP (APS MAX)  predicted Fantasy Points for Average Predicted Saves (MAX, +0.5 for each)
>1.5CG (%) — chances to concede more than 1.5 goals 
FP (CG) — predicted Fantasy Points for possbile conceded 1.5+ goals (-1 for each 2 CG)
WIN (%) — clean Sheet odds (+0.3 for win)
FP (WIN) — predicted Fantasy Points for possbile victory (+0.3)

Σ PFP MIN — a minimum sum of predicted Fantasy Points for goalkeeper
Σ PFP MAX — a maximum sum of predicted Fantasy Points for goalkeeper

These parameters include +2 points for 60+ minutes played as well. Please, remember that these numbers show us the probabilities for good/bad performances rather than exact fantasy points. 

Value MIN — ΣPFP MIN / Price
Value MAX — ΣPFP MAX / Price

My Picks

Avoid

Ederson, Mendy, Guaita, Ramdsale, Alisson. It's always better to pick defenders in such cases.

Schmeichel. Everton won't make a lot of shots on target and The Foxes will be exhausted after their Thursday's match.

Forster & Foster. The Saints keep a clean sheet very rarely on away turf. There were just 2 clean sheets away from home this season. I can't see why they won't concede this time as well. Watford's chances to keep a clean sheet are close to zero from my point of view.

Good Value 

Brentford kept a clean sheet 4 times in the last 5 home matches. Southampton have an ability to make a lot of shots. 8.2 for Raya looks like a good deal considering their chances to keep a clean sheet (31.7%).

I admit that Everton can park a bus once again because it's obviously working better than they play another way. Everton have kept a clean sheet against Chelsea and their situation with possible relegation is pretty tense. That's why I think 7.1 for Pickford it's another good deal. Moreover, considering Leicester's habit of shooting a lot.

The similar reasons encourage me to pay attention to Pope (7.1).

Best Budget Option

Jose Sa (6.0) is an easy best budget pick for me this week. Meslier (5.6), Lloris (5.5) and Dubravka (5.0) have moderate chances to make good saves, but Sa have best chances to not concede a lot. Wolves are the most defensive team among all of the budget teams this week. They had some problems in the last weeks, but they played most matches with initiative and high possession. This match will be played with an obvious Chelsea's possession advantage and I think we can use it.

If you don't have enough budget, feel free to use Lloris, Dubravka and Meslier. (in this order). I believe they can offer us good value, but don't expect a lot.

Montafly
Author

Montafly

Matvey, also known as Montafly is a big fantasy fan, Excel master and music lover who likes making useful spreadsheets and digging up interesting statistics for fantasy games. He has been playing DFS since 2019 on FanTeam and for 10 years on FPL. FanTeam's 2019 Winter Championship of Fantasy Football finalist. FanTeam's 2021 Spring Championship of Fantasy Football finalist and prize-winner. 38th in FanTeam EURO 2020 £1M Game

Comments