23 September 2013

The whole picture: Goals scored not only indicator of late-game success

Late heroics helped San Jose to the 2012 Supporters' Shield, but
two late goals were not enough in their playoff defeat.
I have the privilege of filling in on the Comcast SportsNet New England broadcast of Saturday's New England Revolution vs. Houston Dynamo game, so I've been preparing as much as possible. That involves watching both teams - recent games and past highlights - and studying the numbers.

One number that jumped out as a positive for both teams was their performances in the last 15 minutes of games; both New England and Houston have scored 13 goals this year from minutes 76-90. The league's stats and standings release, which comes out after every gameday, includes two separate tables detailing goals scored and allowed, by each team, broken down by 15-minute intervals. What is not directly released (and often not publicized), however, is teams' goal differential during different segments of the game. That, to me, is the more important statistic. After all, what good is scoring a bunch of late goals if you're allowing even more?

TeamLast 15 minsOverall
Real Salt Lake1912+7+14
Kansas City105+5+15
New England139+4+8
San Jose106+4-10
New York1614+2+11
FC Dallas1414E-2
LA Galaxy1218-6+10
Toronto FC614-8-19
D.C. United311-8-29
Chivas USA622-16-25
So I set out to see how Houston and New England ranked within MLS as late-game success stories, and the numbers I found across the league surprised me in a few ways. Check out the table at right, which lists teams' goals for and goals against in the final 15 minutes, as well as how that goal differential compares to their overall goal differential.

I am not going to try to identify all general trends in this data; for one thing, it's too small a sample size. But I see some really interesting notes about different teams:
  • Houston, Philadelphia, San Jose, and Dallas are the only teams whose goal differentials in the final 15 minutes are BETTER than their overall GD.
  • Incredible efforts by Seattle (2 goals allowed) and Portland (4) in terms of late defense. It's very hard to keep that up over a whole season, and since I think of defense as tending to be more consistent than offense, I think it bodes well for their postseason chances. Obviously, Houston, KC, and Colorado (5) are right behind.
  • I would not have guessed that the LA Galaxy would be -6 in the final 15 minutes and have the league's 2nd-worst late defense, better only than Chivas USA. Yet they still have that +10 overall goal differential.
  • Seeing Seattle, Real Salt Lake, Portland, and Kansas City all in the top five in late goal differential just seems to make sense, since all four feel like quality MLS Cup contenders.
  • The identical -6 late goal differentials of Montreal and LA make me a little less confident on their chances for playoff success.
  • Interesting to note the contrast in late-game fortune for strugglers D.C. United and Chivas USA. D.C. is -21 in the first 75' of a game (presumably, the result in D.C. games is usually decided before the last 15'), while Chivas is only -9 in the first 75' and tends to get punished late.

The most interesting question this raises to me is: Are the final 15 minutes a better or worse indicator of how good a team is? In some respects, you should put more stock in performance over the first 75 minutes than the last 15, because it's a larger sample size. But at the same time, the ability to pull out a late result and come through "in the clutch," to use a sports cliche, is one that is highly valued.

In MLS circles, one has to think of last year's San Jose Earthquakes, who scored 20 goals from the 76th minute on en route to the Supporters' Shield title, which was widely considered an over-achievement. They promptly fell in the playoffs, going a damning -3 in the first 75' of games in their two-leg playoff series, while only compensating with 2 goals in the final 15'.

What this means for New England vs. Houston
Is Houston's -7 goal differential in the first 75 minutes of games this year a sign that the Dynamo are just not that good? Perhaps they will regress to the mean, allow some late goals, and fail to reach the postseason. Or is their +8 goal differential in the final 15 minutes an indicator that they save their best for when it counts the most, as seen in their recent playoff runs? Perhaps they will make another deep run. Or can the difference be explained by the suffocating Houston heat, which wears down opponents and allows Houston to pounce late in home games? That actually is not the case, as Houston is +9 at home and -8 on the road, but in the final 15 minutes, the Dynamo are +4 both at home (6-2) and on the road (7-3).

New England, meanwhile, is a more conventional +4 at home (8-4) and even on the road (5-5), but it should be noted that the Revolution began the year allowing six late goals, yet they are now +10 in the final 15 minutes since losing at home to Real Salt Lake on May 8 (both RSL goals came in the final 15). The goal that turned the tide was Juan Agudelo's 84th-minute tally to clinch a 2-0 win at Houston on May 18. The Dynamo returned the favor with a 79th-minute game-winner from Adam Moffat at New England on July 13, though Moffat is no longer with the club. So maybe we're in for yet another late goal on Saturday night!

This list creates more questions than answers, which is part of the fun. But my larger point is that I hope media and fans factor in late goals scored AND allowed when considering a team's late fortunes.

17 September 2013

Blame it on the rain (well, actually, blame it on the lightning)

When traveling to work on a broadcast, things are usually pretty simple. Fly in the day before, stay two nights in a hotel, and head home the day after the game. But things were a little different for Saturday’s FC Dallas-Colorado Rapids broadcast, and it almost got me in trouble.

Weather delays -- not part of the plan
You see, I am currently in the middle of moving to New York City, a big deal on the personal side that probably deserves several posts of its own. Since I need to use as much of my time as possible to look for an affordable and passable place to live, my trip to Denver was a whirlwind that involved flying out that morning and flying out that night. What nobody had counted on was the Colorado weather, which almost derailed those carefully laid plans.

My outbound trip was smooth enough, catching an early morning flight out of LaGuardia Airport, connecting in Dallas-Fort Worth (where I ran into an old friend from Rice University), and arriving in Denver before noon. Even a little construction on I-70 didn’t keep me from getting to the hotel (kindly reserved even though I had no intention of spending the night), getting lunch with broadcast partner Brian Dunseth, and catching up with two close friends from college.

But the skies turned threatening in the late afternoon, and as I drove to Dick’s Sporting Goods Park, Dunny and I found ourselves plotting how we could stay as dry as possible while running through a downpour to the stadium. The stadium exit was closed off due to flooding, so we had to U-turn and find our way back. Thankfully, the downpour had not yet reached the stadium, so we made it to the press box without having to dance between the raindrops.

Pretty soon, however, lightning started to flash in the distance, and that’s when things got nervy. Any lightning strike in the area requires at least a 30-minute wait before the teams take the field. Soon the possibility of a delayed kickoff came up, and that was an issue for me.

Knowing that soccer games reliably last two hours, I had assumed that I had plenty of time from the game ending (9 p.m. MT) to my flight’s departure (1 a.m. MT) to go back to the hotel, change, return my rental car, and check in for the flight. But a delay, especially one that threatened to go as long as two hours, could have made that impossible. There was also the worrying possibility that the game would be canceled, and that seemed to mean arranging a new flight and calling the game the next morning without a partner (Brian had other Sunday commitments).

Kickoff was initially pushed back to 7:52 p.m. MT, but there was strong speculation in the broadcast booth that it might not actually get underway until 8:30. And that wasn’t even factoring in the possibility of an in-game delay. So I was sweating the timing just a bit.

Once we taped our open, all went smoothly.
We went on-air briefly at 7 p.m. to let TXA 21 viewers back in Dallas know that the game had been delayed, and then we sat and waited. We taped the opening segment of our show a few minutes later, as the players warmed up in the rain, and thankfully no more lightning appeared. In fact, the rain itself had stopped by kickoff, and the game was played on a slick but safe field, and a surprisingly healthy crowd stuck around to see it.

Reunited with Brian Dunseth.
I felt like we ended up with a very good broadcast, discussing both teams’ seasons when the game had a little less action, and talking about tactical changes and the implications of the game once it kicked into high gear. Unfortunately for our viewers back in Dallas, FCD let a 1-goal lead slip away, and Colorado took a big step toward the post season.

Fortunately for me, the game ended on time, and while I drove pretty aggressively on my way out of the stadium and then from the hotel to the airport, I really had plenty of time. Even a wait for my rental-car bus and the crazy Arizona State-Wisconsin ending couldn’t keep me from making my flight and arriving at New York’s JFK Airport early Sunday morning.
Given the conditions in Colorado,
13,145 was a great crowd.

In the future, however, I think I’d like to keep things simple.