2018 - Running With the Numbers and the Quest for Everything

I don't think the Dome thing is a factor separate from the H/A issue which I already discussed and adjusted for. The team gets 2.82 PPG at Home and 1.00 Away. Dome is 4-0-0 at Home and 0-1-0 Away. If you adjust or both H/A and PV/Dome, you're adjusting for the same thing twice I think.

But as I initially wrote, I do agree the W/L is not the whole story: opponents, the rest of the lineup, and other factors you mentioned are relevant. I also explicitly did not claim that the W/L record proves the team is better without Villa.

But, some things you mention, like whether the team relaxes when he starts and steps up when he's out -- those do not disprove the theorem either. They just maybe show it isn't Villa's fault. The team can be better without him without him being to blame.

Finally, circling back to Dome and H/A, this reminds me to mention that I think people make a mistake when they say the club is better under Dome because we won 3 in a row after a long period of uneven results. I think we're better because of the eye test. But in terms of results, all Dome has done is not lose a home game. PV was just as good at that.

I think the Dome difference are the in game adjustments we’ve seen. And also willingness to not play purely out of the back. I’m not sure PV goes 3-0 at home vs those opponents and sticking to his one and only strategy.
 
I think the Dome difference are the in game adjustments we’ve seen. And also willingness to not play purely out of the back. I’m not sure PV goes 3-0 at home vs those opponents and sticking to his one and only strategy.
We lose to RB badly if PV was in charge and we had a depleted roster.
 
Good stuff, as usual.

Will be interesting to see the home/away splits and remaining games for the contenders. Might get on that if I have time.

Small correction to the East line graph - Atlanta are on 41 points, not 40 (as much as it seemed that they got nothing out of the Seattle game).
 
We lose to RB badly if PV was in charge and we had a depleted roster.
The difference between Dome and PV in this scenario is Dome has worked in CL and EL situations his entire career, so he has experience on squad rotation and how to see which players are exhausted or not. PV, while he may have played in these games, did not coach them or have to keep track of which players were tired, etc. So I think that's another reason why we were able to survive a 3-gamer.
 
The team needs to start getting results on the road - and by results I mean wins. It is awesome that NYCFC is lights out at home, and should really be favored in every remaining home game, but until they can turn it around and win road games they don’t have a shot at the Shield or CCL. Atlanta has 20 (!!) road points, RBNJ has 13, NYCFC has 9. With four out of the next five on the road it’s time to start getting it done.
 
The difference between Dome and PV in this scenario is Dome has worked in CL and EL situations his entire career, so he has experience on squad rotation and how to see which players are exhausted or not. PV, while he may have played in these games, did not coach them or have to keep track of which players were tired, etc. So I think that's another reason why we were able to survive a 3-gamer.
Good catch. Thanks. I’ll fix it some time today or tonight.
 
  • Like
Reactions: Shwafta
Good stuff, as usual.

Will be interesting to see the home/away splits and remaining games for the contenders. Might get on that if I have time.

Small correction to the East line graph - Atlanta are on 41 points, not 40 (as much as it seemed that they got nothing out of the Seattle game).
I've actually started putting this together over the weekend for us, Atlanta, and NJ and compared each team's PPG for H/A along with their opponents. I'm going to put up a post a bit later summarizing this.
 
  • Like
Reactions: Gotham Gator
I've actually started putting this together over the weekend for us, Atlanta, and NJ and compared each team's PPG for H/A along with their opponents. I'm going to put up a post a bit later summarizing this.

Great stuff.

Looking quickly at our road schedule.
  • We have played 9 road games - 6 against teams above the playoff line (1-2-3) and 3 against teams below the playoff line (1-2-0).
  • We have 8 road games remaining - 2 against teams above the playoff line and 6 against teams below the playoff line.
 
The difference between Dome and PV in this scenario is Dome has worked in CL and EL situations his entire career, so he has experience on squad rotation and how to see which players are exhausted or not. PV, while he may have played in these games, did not coach them or have to keep track of which players were tired, etc. So I think that's another reason why we were able to survive a 3-gamer.
Yeah perhaps, but I was focusing more on PV’s stubbornness to play out of the back at all costs versus RB.
 
  • Like
Reactions: Jock and Shwafta
Good stuff, as usual.

Will be interesting to see the home/away splits and remaining games for the contenders. Might get on that if I have time.

Small correction to the East line graph - Atlanta are on 41 points, not 40 (as much as it seemed that they got nothing out of the Seattle game).
So I looked at the remaining schedules for us, NJ, and Atlanta, along with each of our H/A PPG and our opponents' H/A PPG. To get a sense for who had the easiest remaining schedule based on their H/A form and opponents' H/A form, I subtracted the opponents' H/A PPG from ours (and NJ/Atlanta) H/A PPG to arrive at a net PPG that I then totaled.

(Example: our road PPG is 1.00, and later we travel to Columbus, whose Home PPG is 1.91. That gets us a net -0.91 for that match)

It isn't scientific, and I'm sure it has its flaws, but that's all I got for now (open to suggestions to make this better).

Totaling up the remaining games, I arrived at Atlanta having the easiest schedule with a 9.63, followed by NJ at a 8.18, and then us at 7.26.

In this analysis, what hurts us the most is our poor road form and our remaining unbalanced H/A schedule. Whereas Atlanta is earning 2.00 PPG on the road and they have road trips to Orlando (1.60 HPPG), Colorado (1.09 HPPG), TFC (1.22 HPPG), and San Jose (0.75 HPPG) on their schedule. Their remaining home games are also all against teams with 1.00 RPPG or less (with two at 0.50 or less).
 
So I looked at the remaining schedules for us, NJ, and Atlanta, along with each of our H/A PPG and our opponents' H/A PPG. To get a sense for who had the easiest remaining schedule based on their H/A form and opponents' H/A form, I subtracted the opponents' H/A PPG from ours (and NJ/Atlanta) H/A PPG to arrive at a net PPG that I then totaled.

(Example: our road PPG is 1.00, and later we travel to Columbus, whose Home PPG is 1.91. That gets us a net -0.91 for that match)

It isn't scientific, and I'm sure it has its flaws, but that's all I got for now (open to suggestions to make this better).

Totaling up the remaining games, I arrived at Atlanta having the easiest schedule with a 9.63, followed by NJ at a 8.18, and then us at 7.26.

In this analysis, what hurts us the most is our poor road form and our remaining unbalanced H/A schedule. Whereas Atlanta is earning 2.00 PPG on the road and they have road trips to Orlando (1.60 HPPG), Colorado (1.09 HPPG), TFC (1.22 HPPG), and San Jose (0.75 HPPG) on their schedule. Their remaining home games are also all against teams with 1.00 RPPG or less (with two at 0.50 or less).
Totally plausible, but we have a new coach. Based on the these past five games I'd think our performance on the road could possibly be vastly better. Heck, I'd settle for "merely" *somewhat* better.
 
Totally plausible, but we have a new coach. Based on the these past five games I'd think our performance on the road could possibly be vastly better. Heck, I'd settle for "merely" *somewhat* better.
That's the part I enjoy about this now, is I can update this as games occur, so the PPGs will adjust as well the remaining difficulty (we just removed an "easy" home game against Columbus, whereas NJ removed a slightly more tough home game against SKC).
 
That's the part I enjoy about this now, is I can update this as games occur, so the PPGs will adjust as well the remaining difficulty (we just removed an "easy" home game against Columbus, whereas NJ removed a slightly more tough home game against SKC).
Lucky fucks had no business winning that.

I’m at the point now where I hate Jersey because of Red Bull, not vice versa as one might expect.
 
NJ defense looked a bit shaky. They won't get bailed out by those wonder strikes every game.

At least, that's what I'm telling myself
 
So I looked at the remaining schedules for us, NJ, and Atlanta, along with each of our H/A PPG and our opponents' H/A PPG. To get a sense for who had the easiest remaining schedule based on their H/A form and opponents' H/A form, I subtracted the opponents' H/A PPG from ours (and NJ/Atlanta) H/A PPG to arrive at a net PPG that I then totaled.

(Example: our road PPG is 1.00, and later we travel to Columbus, whose Home PPG is 1.91. That gets us a net -0.91 for that match)

It isn't scientific, and I'm sure it has its flaws, but that's all I got for now (open to suggestions to make this better).

Totaling up the remaining games, I arrived at Atlanta having the easiest schedule with a 9.63, followed by NJ at a 8.18, and then us at 7.26.

In this analysis, what hurts us the most is our poor road form and our remaining unbalanced H/A schedule. Whereas Atlanta is earning 2.00 PPG on the road and they have road trips to Orlando (1.60 HPPG), Colorado (1.09 HPPG), TFC (1.22 HPPG), and San Jose (0.75 HPPG) on their schedule. Their remaining home games are also all against teams with 1.00 RPPG or less (with two at 0.50 or less).
Nice.
I wouldn't call it a measure of schedule strength, because each team's own performance is given as much weight as the schedule. Set aside the H/A split for a moment. We get 2.0 PPG and Chicago gets 1.1. Under this measurement it's pretty much impossible for Chicago to have an easier schedule than NYC, even if it is objectively easier. But to be clear, I'm not saying this as a criticism of the method, just saying it's not really a schedule measurement.

It is kind of a blunt instrument. With our H/A split, it probably shows we should win every home game and lose every away game, and we don't need you to go to the trouble of doing all this data collection and math to know that. On the other hand, the relative difference between those numbers for the 3 teams you analyzed does tell us something about the overall picture that we wouldn't otherwise know.

It might be worth adjusting for number of remaining games. Because when you do a simple addition, the fact that Atlanta has 13 games left, 14 for us and 16 for RB means it is not an even comparison. Maybe a simple average? Or average home and away separately, then add those two figures? That's what I would play with.

ETA: My last suggestion is flawed. If you average H/A separately then add them it flattens away any difference between the number of home and road games.
 
Last edited:
Totally plausible, but we have a new coach. Based on the these past five games I'd think our performance on the road could possibly be vastly better. Heck, I'd settle for "merely" *somewhat* better.
I think it would be reasonable to expect better on the road for the remainder of the year even if PV stayed. His road record for 2016, 2017 plus March 2018 was 1.43 PPG. He didn't suddenly forget everything. I do have more confidence in Domé, but I expected the club to win 2-3 more away games this year even if Vieira stayed.
 
Nice.
I wouldn't call it a measure of schedule strength, because each team's own performance is given as much weight as the schedule. Set aside the H/A split for a moment. We get 2.0 PPG and Chicago gets 1.1. Under this measurement it's pretty much impossible for Chicago to have an easier schedule than NYC, even if it is objectively easier. But to be clear, I'm not saying this as a criticism of the method, just saying it's not really a schedule measurement.

It is kind of a blunt instrument. With our H/A split, it probably shows we should win every home game and lose every away game, and we don't need you to go to the trouble of doing all this data collection and math to know that. On the other hand, the relative difference between those numbers for the 3 teams you analyzed does tell us something about the overall picture that we wouldn't otherwise know.

It might be worth adjusting for number of remaining games. Because when you do a simple addition, the fact that Atlanta has 13 games left, 14 for us and 16 for RB means it is not an even comparison. Maybe a simple average? Or average home and away separately, then add those two figures? That's what I would play with.

ETA: My last suggestion is flawed. If you average H/A separately then add them it flattens away any difference between the number of home and road games.
S0 I did average this based on the number of games remaining, and it does have us jump ahead of the Red Bulls. Atlanta has an average PPG difference of 0.74, we are at 0.52, and NJ is at 0.51.

Additionally, some more stats solely based on remaining schedules, not incorporating ours (or NJ/Atlanta's) success:

We have 8 remaining away games against teams that average 1.72 PPG at home (best is DCU at 2.33, this will come down and the worst is Seattle at 1.00, I imagine this will go up a bit), and 6 remaining home games against teams that average 0.95 PPG on the road (Best is NJ at 1.44 and Worst is DCU at 0.58).

NJ has 7 remaining away games against teams that average 1.86 PPG at home (best is NYC at 2.82 and they also have 2 against DCU's inflated HPPG as noted above, Worst is San Jose at 0.75). NJ has 8 remaining home games against teams that average 1.00 PPG on the road (best is Atlanta at 2.00 and worst is TFC at 0.50).

Atlanta has 7 remaining away games against teams that average 1.65 PPG at home (best is NJ at 2.44 and worst is San Jose at 0.75) and have 6 remaining home games against teams that average 0.71 PPG on the road (best is NER at 1.00 and worst is RSL at a dismal 0.40)
 
If we wanted to reach 20, that would be 8 more wins, which would be another 24 points... almost leveling us with Toronto from last year. Chances high? who knows! With 6 home games left, if we can win every one, that's another 6 wins...