geez you guys can't take a joke >_> I'm poking some fun
2018 Season Points Prediction
- Thread starter dannd23
- Start date
54 or less. I'm really pessimistic about Villa remaining healthy all year and not loosing too much to Father Time.
I think Villa is out for maybe a month and slows down his prolific scoring, and in true NYCFC fashion no one else can really make up the numbers.
Pre season pessimism ahoy.
I think Villa is out for maybe a month and slows down his prolific scoring, and in true NYCFC fashion no one else can really make up the numbers.
Pre season pessimism ahoy.
D
dummyrun
Guest
Their rankings and predictions don't account for offseason roster changes. FiveThirtyEight's got a great model but it's pretty much useless after the biggest offseason in MLS history until they feed some results into it.
Their model is not built for MLS. Not even close.
They’re going to pass the ball to Mata on the deep left sideline, and then sit back and watch.
Seven
Registered
Our defense is going to make bad passes and bonehead decisions. It's just the nature of our system of playing it out of the back. I absolutely think we have gotten better at it, but those brain farts are still going to happen.
I'm sure we will also get caught on the break a few times, especially with both full backs wanting to bomb forward
D
dummyrun
Guest
Why?
Last edited:
I tend to think folks are right that it doesn’t account for offseason moves and over-weights prior season results. That said, I’m not sure what else you could do.
Bottom line is any model is going to struggle as the distribution of outcomes becomes more tightly grouped.
Still, I’d rather have something to look at over nothing. So c’est la vie and shit.
D
dummyrun
Guest
I'm not sure those problems are unique to MLS, except maybe this offseason since rule changes led to more turnover than usual.
What else you could do: FiveThirtyEight's model already relies on Transfermarkt values (which are sketchy but better than nothing) in its league coefficient. You could use the change in an individual squad's value as a proxy for how much better or worse it got in the offseason. Maybe it'd be more predictive than just looking at last season's performance, maybe not.
If you built Transfermarkt squad values into the model, it would probably show NYCFC losing ground to other top teams in the East this offseason. I'm not sure that's true. But that doesn't necessarily mean it'd be a bad way to do preseason rankings—it would just mean we played moneyball while Atlanta was busy buying the entire continent of South America.
dannd23
Registered
Yeah, no doubt about all of this. I was more being hyperbolic because on paper, our team is really freaking good even without Villa, and most of MLS is not.
IIRC, they account for the quality of the new personnel a team acquires based on transfer fees paid. That is a good proxy in the EPL, but a very poor proxy in MLS, where teams frequently bring in very talented players without paying a transfer fee.
D
dummyrun
Guest
They use Transfrmarket for rating a league's strength. For a team's strength, they do rely on transfer fees, at least to start a season. You have to click through your link to the one on their 4 metrics to find the following.
As with our Elo-based rating systems, each team’s ratings change in the offseason. Rather than reverting each team toward the same mean, we revert it toward a time-weighted average of its final rating over the past five seasons. In addition, we adjust each team’s preseason rating based on players it acquires or sells in the offseason.
https://fivethirtyeight.com/features/how-our-club-soccer-projections-work/#fn-7
Specifically, these adjustments are based on subtracting transfer fees a team got in the offseason from how much it spent on acquiring players, relative to league average. For every standard deviation of net spend above league average, a team’s rating is boosted by about 0.09 points, split evenly between the team’s offensive and defensive ratings.*
https://fivethirtyeight.com/features/how-our-club-soccer-projections-work/#fn-7
* - Specifically, these adjustments are based on subtracting transfer fees a team got in the offseason from how much it spent on acquiring players, relative to league average. For every standard deviation of net spend above league average, a team’s rating is boosted by about 0.09 points, split evenly between the team’s offensive and defensive ratings.
https://fivethirtyeight.com/features/how-our-club-soccer-projections-work/#metricshttps://fivethirtyeight.com/features/how-our-club-soccer-projections-work/#fn-7
* - Specifically, these adjustments are based on subtracting transfer fees a team got in the offseason from how much it spent on acquiring players, relative to league average. For every standard deviation of net spend above league average, a team’s rating is boosted by about 0.09 points, split evenly between the team’s offensive and defensive ratings.
D
dummyrun
Guest
Nice catch, thanks for pointing this out. Bizarre that a team's average performance over the last five seasons might affect its preseason ranking more than its offseason roster moves, but I guess the stats nerds have their reasons.
If you compared our expected goals performance and our actual goals performance last season against our rating, it made no sense - at least not until the end of the season.
Because even the best, most data-intensive modeling of past performance and new additions and other changes will never be more accurate than just assuming Arsenal will finish 4th.
Similar threads
- Poll
