When I wrote down Lloyd, I forgot Lloyd was out on a red card. But I won't revise the score prediction, but will attribute the SB goal to McCaskill, not Lloyd. - even with Shea Groom out with concussion. ...OTOH.........
Lunatica is on a great roll, has moved into second place, and may be closing on the robot! UTBalance isn't too shabby either. PS - The robot's operator was very happy for the robot to take a 0 on the Pride v Thorns game.
For my results, I just gotta make the same post-game quote from every coach who sees a playoff spot slipping away: "To get at least a point from every game is a good thing." In the break between U20 England v. Mexico and Netherlands v. France, I did some arithmetic (errors may abound!). Since the end of Week 14 (July 7), Lunatica has gained 47 points and Blue Crimson has gained 44 points. Others beating the robot’s 35 points are Tracer with 37, BalanceUT with 36, and me with 36.
I've noticed that, and I have a theory: The humans are learning from the robot and the other statistics. I'm not sure the theory is right. If I have time, I'll do another week-by-week cumulative score chart that shows how predictors have trended over the course of the season. I'm guessing it will show that the robot's performance has not tailed off but rather has been pretty consistent. Rather, the humans have improved. If my guess is right, the question is why. My proposed answer will be the humans are learning from the robot and the other stats -- as they should. Hopefully, someone (or more) will pass the robot before the end of the season.
Interesting theory. By my part, I don't use to look at robot's predictions before making mine. On the other hand, this somehow could reinforce the theory, since I am one of the humans who definitely aren't catching up with the robot... But probably you mean something more wide than just looking at robot predictions: you mean "learning to use its same methods", although I am not sure what this could exactly mean. You mean that we'd start actually using the data from the previous matches of the season (or using them better) to make more consistent predictions? I don't think, though, that humans (at least consciouly) have started applying this kind of statistical methods in a more rigorous way and suddenly forgot all of the other elements, including their personal bias. Maybe it's just a question that, towards the end of the season, the relative strengths of the teams are so much apparent that humans can have a solid grasp on them? But this shouldn't make their predictions "as good" as the robot's ones, instead of "even better"? Or maybe it's possible that, by the end of the season, motivations of the teams (who can now see their "clock" for respective target closing down) become somehow more relevant than statistical history (mind, just another theory, with no actual data to support it ).
You could test the theory by posting the robot predictions after the week’s guessing is done and see if the human predictions degrade.
Predictions for all of this week's games except Portland v Chicago, which will await the result of Wednesday's game: Seattle 1 v Chicago 1 Houston 1 v Washington 1 NJNY 1 v Utah 2 North Carolina 1 v Orlando 1
Here's a chart that shows predictors' scoring trends over the course of the season. Some weeks have more games than others which has some effect on the chart. For example, week 10 is relatively flat -- because it had only one game. As I suspected, the robot is relatively consistent from week to week -- more or less a straight line progression. On the other hand, Lunatica and Blue Crimson have improved in the relatively recent period.
Wimpy prediction - I can't lose ground. SEA 1 CHI 1; so far only @Tracer and @BlueCrimson have bucked the tide, they both could possibly gain on the robot.
Actually not!!! But I won't change my prediction anyway and will cross my fingers. Also, by reading the injury report, I discover Yanez has a foot fracture, that I wasn't aware of. Meanwhile, Utsugi's calf strain looks more serious than we had initially thought: I hope we're going to see her back sooner than later. And Yaverbuch never played in the whole season.
I'll have to ask the robot what it thinks about crossing fingers. I think it will say, "Duh, what does that mean?"
I already picked for tonight's game, but here are my calls for the weekend: HOU 2 WAS 1 NJ 0 UT 1 NC 3 ORL 1 POR 1 CHI 1
Fourth game prediction for the weekend: Portland 2 v Chicago 1 (could be 1-1, but the robot's protocols say not as likely)