:format(webp)/cdn.vox-cdn.com/uploads/chorus_image/image/51907623/usa-today-9266443.0.jpeg)
Heading into the 2016 MLS season, not a lot of people had the Colorado Rapids on their "teams to watch" list. Much of this can be attributed to their 2015 season that saw the team earn a Western Conference worst 37 points while scoring the fewest goals in the league. Managerial maestro Pablo Mastroeni, however, has maneuvered his outfit to a near worst-to-first turnaround, missing out on the Supporter's Shield by a mere 2 points.
Much of their success this season came down to two factors: a staunch defense and home form. Offense was a problem, but with their defense being so stout the Rapids didn't need to score often. They proved this by winning a league-high eight 1-0 matches to go with a league-high 13 shutouts.
When it comes to analyzing the Audi 2016 MLS Cup Western Conference Finals matchup between these Rapids and the Seattle Sounders, however, it's a little shallow to say "Colorado defense good, offense bad. Seattle defense okay, offense okayer," so I wanted to dig a little deeper.
Earlier this season I wrote a story looking at the luck-vs-unluck of the Sounders season through 23 games, using American Soccer Analysis and their Expected Goals (xG) models. What I found was that the Sounders were mostly unlucky, underperforming their Expected Goals For (xGF) by quite a goodly sum, which (unsurprisingly) affected their points as a result. Using this same methodology, I wanted to look at the Colorado Rapid's season to determine their relative luck or unluck.
A quick caveat: these numbers are only through the first 33 games, excluding the Decision Day matchups (please update your site, ASA). In their final game, a home match versus San Jose, Colorado drew 1-1.
Since the first leg of these Western Finals are being played in Seattle, I first wanted to see how Colorado performed on the road.
| Road Opponent | GA | xGA | GF | xGF | xGD | Points | xPoints |
| San Jose Earthquakes | 1 | 0.51 | 0 | 1.28 | 0.77 | 0 | 3 |
| DC United | 1 | 2.26 | 1 | 1.00 | -1.26 | 1 | 0 |
| Real Salt Lake | 1 | 0.99 | 0 | 0.83 | -0.16 | 0 | 1 |
| Sporting Kansas City | 1 | 2.16 | 2 | 0.77 | -1.39 | 3 | 0 |
| Montreal Impact | 2 | 0.94 | 2 | 1.29 | 0.35 | 1 | 1 |
| Columbus Crew | 1 | 1.48 | 1 | 1.00 | -0.48 | 1 | 1 |
| Seattle Sounders | 0 | 1.11 | 1 | 0.75 | -0.36 | 3 | 1 |
| LA Galaxy | 0 | 1.10 | 0 | 0.95 | -0.15 | 1 | 1 |
| Vancouver Whitecaps | 2 | 1.80 | 2 | 1.34 | -0.46 | 1 | 1 |
| NYC FC | 5 | 2.65 | 1 | 0.45 | -2.20 | 0 | 0 |
| LA Galaxy | 1 | 2.51 | 1 | 0.77 | -1.74 | 1 | 0 |
| Real Salt Lake | 2 | 3.23 | 1 | 0.40 | -2.83 | 0 | 0 |
| New England Revolution | 2 | 1.59 | 0 | 1.32 | -0.27 | 0 | 1 |
| FC Dallas | 0 | 1.20 | 1 | 0.37 | -0.83 | 3 | 0 |
| Vancouver Whitecaps | 3 | 1.73 | 3 | 1.25 | -0.48 | 1 | 1 |
| Houston Dynamo | 2 | 0.63 | 3 | 0.59 | -0.04 | 3 | 1 |
| Portland Timbers | 1 | 2.89 | 0 | 0.69 | -2.20 | 0 | 0 |
| Totals | 25 | 28.78 | 19 | 15.05 | -13.73 | 19 | 12 |
The Rapids earned 19 points from a road record of 4-6-7 (W-L-T), good for 1.12 PPG, and in the process accumulated a GD of -6. However, according to their xGD, the Rapids should have earned only 12 points on the road while accumulating a GD of -13.73 (a quick methodology note: any xGD between half of a goal either way I count as a draw, anything greater than half of a goal difference is a loss or a win) .
On both defense and offense the Rapids over-performed. They scored 3.95 goals greater than their xGF predicted, while simultaneously allowing 3.78 goals fewer than their xGA predicted. This is how they were able to only have a road GD of -6, with a majority of that being from the road game against NYCFC.
A couple notes:
- In only two road games did Colorado have a better xGF than their opponent.
- Colorado only had 3 of their 13 shutouts on the road.
- In only seven games did Colorado have an xGF of 1.00 or greater.
- In only four games did Colorado hold their opponent to an xGF of less than 1.00.
- In 10 road games against other playoff teams, Colorado had an xGD of -10.57.
Now for the Home Games, where the second leg of this series will be played.
| Home Opponent | GA | xGA | GF | xGF | xGD | Points | xPoints |
| LA Galaxy | 0 | 0.82 | 1 | 0.76 | -0.06 | 3 | 1 |
| Toronto FC | 0 | 0.64 | 1 | 1.76 | 1.12 | 3 | 3 |
| New York RB | 1 | 0.85 | 2 | 1.02 | 0.17 | 3 | 1 |
| Seattle Sounders | 1 | 0.60 | 3 | 2.82 | 2.22 | 3 | 3 |
| Real Salt Lake | 0 | 1.03 | 1 | 1.34 | 0.31 | 3 | 1 |
| Sporting Kansas City | 0 | 0.45 | 1 | 1.15 | 0.70 | 3 | 3 |
| Philadelphia Union | 1 | 0.50 | 1 | 1.12 | 0.62 | 1 | 3 |
| Chicago Fire | 1 | 0.54 | 2 | 0.88 | 0.34 | 3 | 1 |
| Portland Timbers | 0 | 0.79 | 0 | 0.98 | 0.19 | 1 | 1 |
| Sporting Kansas City | 0 | 0.78 | 1 | 0.68 | -0.10 | 3 | 1 |
| FC Dallas | 1 | 0.64 | 1 | 0.93 | 0.29 | 1 | 1 |
| Vancouver Whitecaps | 0 | 0.26 | 2 | 2.49 | 2.23 | 3 | 3 |
| Orlando Cit | 0 | 0.30 | 0 | 1.00 | 0.70 | 1 | 3 |
| San Jose Earthquakes | 0 | 0.83 | 0 | 1.50 | 0.67 | 1 | 3 |
| Portland Timbers |
0 | 0.38 | 1 | 0.52 | 0.14 | 3 | 1 |
| San Jose Earthquakes | 1 | 1.99 | 2 | 1.74 | -0.25 | 3 | 1 |
| Houston Dynamo | 1 | - | 1 | - | - | 1 | - |
| Totals | 7 | 11.40 | 20 | 20.69 | 9.29 | 39 | 30 |
For their home numbers (again, we're excluding the season finale 1-1 draw against Houston), Colorado was much better than they were on the road*, and again outperformed their predicted xG metrics defensively, but underperformed on offense.
*Award winning analysis
In 16 home games, Colorado went 11-0-5, good for 2.38 PPG, but according to their xGD they should only be 7-0-9, for 1.88 PPG with a xGD of +9.29 instead of the +13 GD they accumulated.
Combining this xGD-determined W-L-T, through 33 games the Rapids should have been 8-7-18, for 42 points and -4.44 GD, instead of the 15-6-12 +7 they ended up with prior to Decision Day.
A couple home notes:
- In only two matches did the Rapids beat their opponents by more than 1 goal (their only two multi-goal wins of the season).
- In 13 of 16 matches Colorado had an xGF greater than their opponent.
- In 10 of 16 matches Colorado had an xGF greater than 1.00.
- In 14 of 16 matches Colorado held their opponent to an xGF of less than 1.00.
- In their 8 matches against Playoff teams, Colorado had an xGD of 4.96, with almost half of that number coming from the Seattle debacle.
It should be noted that perhaps these large variances in xGA & xGF versus actual GA & GF is by design, that it's something Colorado attempts every season, but to combat that I'll briefly look at other seasons where Mastroeni coached the Rapids.
Rapids GD-xGD under Mastroeni
2016: 11.46
2015: -0.53
2014: -19.04
Okay nope. Wow, the Rapids were really unlucky in 2014.
Make no mistake, this 2016 version of Colorado is formidable at home, with a GAA through 18 games of 0.39 and a GFA of 1.17, with an xGAA of 0.71 (through 16 matches) and an xGFA of 1.29. If the Sounders want to advance to the MLS Finals, their best chance is to get a favorable result in Seattle, where Colorado has a road GAA of 1.44 (xGAA of 1.80) and a GFA of 1.06 (xGFA of 0.94).
If Seattle can finish Tuesday's match with a multiple goal lead, Colorado would need to buck their trend of close home games in order to progress. But with only two losses all season of greater than a goal, and only three predicted by xG, beating Colorado by multiple goals has been a tough ask all season.