Math Formulas (via trindade.joao)
The real power rankings should be released anon, but for now it's time to get our crunch on again, ranking teams by what the stats say they're doing, not by stupid standards like how much they're actually winning. But seriously, this is actually a good time of the season to compare the two. Now that we're just over halfway, you'll find more elements of human bias creeping into most power rankings. One of those is that people tend to stop paying attention to teams at the bottom of the table and just coast their ranking from week to week until and unless the team suddenly shoots up the table into playoff contention.
Take Toronto FC for instance. One of the worst teams in the league, right? Maybe the worst? Can't win to save their lives? Rank them 19th? Maybe 18th if you're feeling generous. Well, actually they've only lost 1 game in their last NINE matches. Their 1.8 PPG over that span would give them 60 points over the course of a full season and would put them behind only New York in the current standings. Of course that early season run can't be forgotten, but if power rankings are to any extent a measure of how teams are doing lately, could you seriously rank them even in the bottom 5? Well, all of the current power rankings out there are doing just that. Advantage: crunchy power rankings.
Here's the full table with more discussion below the fold:
|Team||Net Duels||Net Shots
|Sporting Kansas City||2.58||0.84||6.53||80
|New England Revolution||1.02||1.49||0.61||80
|San Jose Earthquakes||0.49||1.26||1.58||77
|Real Salt Lake||-2.35||1.1||4.88||76
|Seattle Sounders FC||2.12||0.86||0.83||71
|New York Red Bulls||1.21||0.26||2||61
|Los Angeles Galaxy||3.4||0.13||-1.4||56
The biggest gainer is the Houston Dynamo at +19 points. It's a team that's quietly risen up the Eastern Conference standings and now find themselves in a top 3 playoff position. These stats suggest that it's no fluke and they're a team destined for the playoffs at their current performance level. The biggest loser is the Portland Timbers, despite the fact that they were already third-to-last the last time we ran the numbers. Fortunately, at 8 points there's really not much more distance for them to fall. Despite (or maybe 'because of') jettisoning John Spencer, the team is in a total freefall, capped with a 5-0 loss to one of the worst teams in the league.
Speaking of FC Dallas, the recency weighting of the CPR give them a big boost after that big win over the Timbers with +12 points and a jump into 12th place. That might be too much of a boost and you could argue that we should cap the impact of a single game on the results. On the other hand, the return of David Ferreira legitimately makes this a much better team, and perhaps 12th best isn't entirely out of line.
The biggest differentiation with the standings are in New England, D.C., and Vancouver. The Revolution are a team that the CPR continue to rate very highly, but who are way out of playoff position in the East. Their recent run of 5 games unbeaten looked to justify that ranking, but they've lost their last two in a row (and to not very good teams). The thesis of the CPR is that their results may not be reflecting their performance and that they may be victims of bad luck. One piece of evidence for that is that they've lost only one match by more than 1 goal over the entire season. Eight total losses by exactly one goal certainly leaves a lot of room for bad breaks and bad bounces to affect a team's results. They're currently the best team in the league in net shots on goal (again, weighted for recency) and over the long term the team that does that is just going to score more goals. But the long term may not be long enough for New England to catch back up and get into the playoffs.
At the other end, both D.C. United and Vancouver are teams that are doing well in the standings but that the CPR just doesn't like. For United, I believe one factor is their relative lack of draws. They score a lot of goals (second only to New York in the East) and they give up a good chunk of goals for a playoff team. High scoring games lead to fewer draws and basic math says that a win and a loss is better than two draws, so over the long term teams that score a lot of goals but draw less are going to get a good share of points. Because I correlate stats with PPG, I don't factor in draw rates directly. And it may turn out that being a +1 net shot on goal is much better if you're averaging 2.5 goals to 1.5 goal against versus averaging 1.5 goals to 0.5 goals against (see: Chivas). Something to look at long term. Then again, DC has lost 3 of their last 5, so maybe they're just not as good as they've looked through the first half of the season.
Vancouver doesn't have a particularly low draw rate. They're just statistically really bad with one of the worst net shots on goal in the league and one of the worst net offensive passing percentages. And that stat isn't going to improve after they shipped out Davide Chiumiento who was one of the better offensive players on the ball in the league. That they're currently in a top 3 playoff spot (ignoring games in hand) despite a negative goal differential suggests that there's some good fortune and/or good timing going on there as well. But if they end the season with a bottom 5 net shots on goal, I'd be shocked if they're in the top three, and I'd even consider their playoff position tenuous if the current non-playoff teams in the West didn't look so abject.