Week 16 Review
Woooooooo! Welcome back. Fall season is my favorite because I can really start to leverage the data from spring season to find those pockets of value, efficiency, and for a nerd like me, fun.
Week 16 put the Behind the Numbers team off to a great start, which is muted only slightly by our Captain picking curse. The Behind the Numbers team put up 119 points for a WR/OR of 54th overall. We also picked up a team value gain of $4.1, which was tied for the best gain I saw. The only thing that hurts is the literal, last second swap from Valeri, who’d been my captain all week, to Fernandez as I attempted to win the weekly/extra time prize. Turns out Valeri was better for that anyway.
We’re to spend some time talking about the value system, how it works, and how to game it. This is in response to so many people saying something to the effect of “man, I scored so many points this week, how did my value not rise more?” For everyone’s reference, the average price gain this week, from people who responded on twitter/reddit, was $2.9 with a std dev of $0.9. So if you scored a gain of between basically $2-$4 you scored in the middle 66% of teams.
First, some data:
This is a plot of what a team scored vs. what that team’s price rise was this week. The line is a linear fit to the data. The key number here is the R^2 value. In this case, it’s saying that only 30% of a team’s price rise is accounted for by how many points they scored last week. It is key rather how the teams go their points, or how the points are distributed between players and how expensive the point getting players were.
Let’s revisit a chart I made very early in the spring season:
This is an estimate I put together of the threshold needed for players at each price point to rise in value a certain amount. The key thing to note here is that for cheaper players, the threshold is lower and vice versa.
Let’s compare two example (real) teams from last week.
Team 1 scored 119 points and got a price rise of 4.1 while Team 2 scored just 1 fewer point, but gained $1.5 less in value. Why is this? Let’s look at how the points are distributed. Team 2 gained a +10 advantage over Team 1 by captaining Valeri, but can only gain a max of $0.5 in value. Meanwhile, look at the players 8, 9 and 10. They make up the 10 points lost due to captain choices but also account for the extra $1.5 in value. The key to gaining value is having all your players score pretty well. This will give a bigger value gain than having some superstars and some duds. This is especially true if the players who score pretty well are dirt cheap.
Looking at my team last week, I gained 2.5 from 5 players priced under $2.5 vs. $1.6 from 6 players priced at $7.5 or higher. So everyone wants to pick cheap players who score well, that’s not surprising. The key is to ask yourself where are the cheap players who can put up good scores. Checking the chart above, for a $5.5 player, you need to score 7 points to get a $0.5 price rise. Now, generally. To make thing simple, you’re probably not going to rack up 2 points for minutes and 5 points from bonus points from any player, let alone someone in the $5.5 range. So we’re looking for $5.5 players who either score goals or get clean sheets. There are guys like Shinyashiki last week who can bag goals to get that price rise or Wondo last season. But far more often and far more predictably, you NEED to be looking at defenders who can bag clean sheets. This is the most predictable path to value gains, and then if they don’t hit, the negative price changes are pretty minor compared to say, Zlatan…happy hunting value seekers.
As a new segment on this column, I’ll be putting together a true DREAM TEAM each week, rather than the canned 4-4-2, team limit defying one MLS puts out. Here are the rules.
- only three players per team
- must be a valid formation
- must fit in the average budget for that week
- ties are broken with lowest value player making the team
- the bench is all 4.0 scrubs
So here’s the Week 16 TRUE Dream Team
Week 17 DGW Team Defense Rankings
Total Budget this Week: Average Budget: $102.9, BTN Budget $104.1
I think there are reasonably 7 teams you could look at GK/DEF from this week; the top 4 + POR + SJE. I just think the data hasn’t caught up to SJE and they’ve got a double home.
Gonzalez (FCD) and Clark (POR) both had good, not great weeks last week. The difference is that Clark is still just $5, meaning that 7 points will still help him going forward into this week.
On the budget end, Brian Rowe (ORL, $5) and Vega (SJE, $5) has pretty good underlying numbers. On the pricier end at $6, Guzan and Hamid have top matchups.
In DGW, I basically just decide to roll with one keeper, no Keeperoo shenanigans unless the DGW guy really tanks and I happen upon some budget.
On DEF, you first need to drop the big hitters from last weekend into your backline. That means FCD and POR since LAG and SKC aren’t playing this week.
Again, we’re looking for the highest point/$. On FCD, that’s Bressanelli at $5.5 and on POR that’s Moreira at $6. Maybe a little pricey, but they’re in that range where they should be guaranteed a price rise this week.
Everywhere else we’re back to hunting for budget where we can since we have no history to go on.
In ATL, your best bet is to watch lineups for who plays the outside back sports. Escobar is listed at red on the site, so there’s potentially Parkhurst at $5.5 + one opportunistic budget option there.
From the other teams, there’s not a ton of value in ORL, but Sane is likely your best option. DC is much the same way, with Jara being the best though expensive option. SJE should offer value with Lima and Cummings out, though that will also cut their backline’s attractiveness. Guys to look at there are Lopez ($4.5) and Affolter ($4.5).
Sadly, I don’t see much other predictable value at the moment. Just keep an eye on Wednesday lineups and take your gambles with cheap guys.
Week 16 DGW Team Offense Rankings
Week 16 Top 50 Player Projections
Note: Player projections extrapolate previous stats and matchup data for the current player and match up assuming that player plays the full 90 (SGW) or 180 (DGW). It does not factor in the probability of a player starting or being rotated during a DGW.
First, let’s look at guys coming off hot weeks. Valeri needs to be in your team with his crazy 21 points from last week. Loria (POR) also has a high point/$. Outside of that, Reyna (VAN, 8.5) is the only guy who’s score can really buoy him this week in the midfield. Along the front line, Ferreira, Badji, Fernandez and Ebobisse are guys to consider based on last week’s performance.
So the rest of our spots can go to guys who should be solid this week. The exception is Pozuelo, who’s sub appearance made him 0.5 cheaper. I think this actually not something to scare us away. We can’t really be looking for $11M guys to give us price value. We just need them to give us points. Enjoy discount Poz!
For stand-alone guys SJE has guys at a good value like Wondo, Espinoza, and Eriksson.
Spring favorites like Monteiro are also still good values.
Carles Gil is a BP machine with two home games and the Revs are showing signs of life.
Rooney and Nani are nice on the production side but aren’t as good a value as some other options.
2019 Week 17 Behind the Numbers Team
Here’s my first draft of the Behind the Numbers Team based on the discussion above.
If you’d like to follow me on twitter, I’m @MLSFB_BTN. I typically post my updated lineup on Friday nights before the games start to show if my thinking has changed since this column went up.’
Resources for nerding out about MLS Fantasy
- http://www.mlscheatsheet.com/ — Bonus point data, projections and more
- https://www.americansocceranalysis.com/asa-xgoals — Interactive xG and passing data
- https://www.whoscored.com/ — Analytics based player ratings