538 now has MLS SPI projections available

[Midas Mulligan]

Traditionalists: "You can't measure soft skills."

Statisticians: "Prove it matters."

I don't know a ton about soccer metrics, but in baseball, the "soft skills" argument was similar to the "god of the gaps." Soft skills were attributed to every contribution that wasn't a hit, run, or RBI. Until those other contributions started getting measured.

Ask any athlete if being "in the zone" is a real thing, and they'll nod emphatically. But then we measured it. Not really a thing.

Not sure how I ended up here, but what is thought now is that your last paragraph is empirically untrue.

 

[mgarbowski]

So somehow, they think we have a better chance of being 1 or 2 in the East, but somehow NYRB has a better chance of being first overall?

It's odd and counterintuitive, to be sure, but I think this is why 538 has this position.
SKC is the most important team in the West for SS purposes, LAFC a bit behind. NYC is a point ahead of SKC but behind on PPG as SKC has a game in hand until this afternoon. SKC controls their destiny vis-a-vis us. RB is way behind both but with many games in hand and tied with us on PPG. In addition to all the extra games RB has a game remaining at home against SKC. They also play LAFC at home. So RB has an opportunity to overtake SKC that we do not, by beating them H2H. Of course if RB wins that home game against SKC we benefit as well, but maybe the algorithm doesn't think that far, or maybe it's just because if RB wins then that's a guaranteed 3 for them while all our other games are still contingent. Also both remaining games between NYC and RB are in NYC so we have the advantage there which is probably why we have better figures for total points, for a bye, and for 1st in the East. The computer shows we have a 47% chance of winning each derby and RB only 30% each.

I'm not saying I agree with this odd inversion you noticed but I do imagine this is the logic behind it.

 

[mgarbowski]

Not sure how I ended up here, but what is thought now is that your last paragraph is empirically untrue.

My reading of the literature of the advanced stats people and also just basic math/stats people is that sbrylski is right. What he said is the majority position.

 

[Midas Mulligan]

My reading of the literature of the advanced stats people and also just basic math/stats people is that sbrylski is right. What he said is the majority position.

https://www.thecut.com/2016/08/how-researchers-discovered-the-basketball-hot-hand.html

 

[mgarbowski]

https://www.thecut.com/2016/08/how-researchers-discovered-the-basketball-hot-hand.html

Thanks. I was not aware of that. I confess the more important aspect of it — the discovery that past random events affect the probability of future random events — bugs the hell out of me, though it seems those who know accept it. Kind of like my initial reaction to the Monty Hall paradox which I now fully accept to the point it is even intuitive to me, but when I was first presented with it I refused to accept it.

 

[mgarbowski]

Also, I’m not sure how much value I put in new “hot hand theory,” which seems to be that once you account for the phenomenon that making 3 shots in a row makes it less likely you make the next, the fact that players perform within 1% of normal means they are actually hot. At the end of the day they are still within 1% of normal, so what do you do with that?

 

[Midas Mulligan]

Thanks. I was not aware of that. I confess the more important aspect of it — the discovery that past random events affect the probability of future random events — bugs the hell out of me, though it seems those who know accept it. Kind of like my initial reaction to the Monty Hall paradox which I now fully accept to the point it is even intuitive to me, but when I was first presented with it I refused to accept it.

I felt about the same, honestly.

 

[adam]

It's odd and counterintuitive, to be sure, but I think this is why 538 has this position.
SKC is the most important team in the West for SS purposes, LAFC a bit behind. NYC is a point ahead of SKC but behind on PPG as SKC has a game in hand until this afternoon. SKC controls their destiny vis-a-vis us. RB is way behind both but with many games in hand and tied with us on PPG. In addition to all the extra games RB has a game remaining at home against SKC. They also play LAFC at home. So RB has an opportunity to overtake SKC that we do not, by beating them H2H. Of course if RB wins that home game against SKC we benefit as well, but maybe the algorithm doesn't think that far, or maybe it's just because if RB wins then that's a guaranteed 3 for them while all our other games are still contingent. Also both remaining games between NYC and RB are in NYC so we have the advantage there which is probably why we have better figures for total points, for a bye, and for 1st in the East. The computer shows we have a 47% chance of winning each derby and RB only 30% each.

I'm not saying I agree with this odd inversion you noticed but I do imagine this is the logic behind it.

They probably also didn’t factor in PV leaving us, Tyler Adams leaving NJ, or for that matter Almiron leaving Atlanta. Nor any possible replacements. The one thing that continues to irk me about MLS, is the 2nd half sometimes becomes a brand new season.

 

[mgarbowski]

They probably also didn’t favor in PV leaving us, Tyler Adams leaving NJ, or for that matter Almiron leaving Atlanta. Nor any possible replacements. The one thing that continues to irk me about MLS, is the 2nd half sometimes becomes a brand new season.

That's all definitely true, but also, I don't think, relevant to the anomaly that Fred noticed.