Lite hockey stats

After Cornell got out to a 1-0 first period lead over Union, it was noted on the broadcast that if Cornell leads after one period, it has thus-and-such outcome, I think 5-1-2. You hear that a lot.

See SBNation on The Value of the First Goal

Eric T, SBNation
Teams that score the first goal win 67 percent of the time. So that first goal really sets the tone, huh? But here's the catch: teams that score the second goal win 68 percent of the time. And teams that score third win 68 percent of the time. So, uh, about that super-important first goal...

Suppose a team wins a game 4-2. They scored 67 percent of the goals. If no goal is any more important than any other, the odds that they scored the first goal will be 67 percent. The odds that they scored the second goal will be 67 percent. And the odds that they scored the third goal will be 67 percent. So you know why teams that score first win two-thirds of the games? Because winning teams score about two-thirds of the goals. There's nothing magical about the first goal, no matter what The Stats Guy might tell you.

There are plenty of other Stats-Lite in other sports. If LeBron has a triple-double, the Cavaliers are 7-3, something like that. We've seen stats that mix cause and effec: Running teams have a better winning percentage. (Because they're ahead comfortably. So they run the ball.)

billhoward
After Cornell got out to a 1-0 first period lead over Union, it was noted on the broadcast that if Cornell leads after one period, it has thus-and-such outcome, I think 5-1-2. You hear that a lot.

See SBNation on The Value of the First Goal

Eric T, SBNation
Teams that score the first goal win 67 percent of the time. So that first goal really sets the tone, huh? But here's the catch: teams that score the second goal win 68 percent of the time. And teams that score third win 68 percent of the time. So, uh, about that super-important first goal...

Suppose a team wins a game 4-2. They scored 67 percent of the goals. If no goal is any more important than any other, the odds that they scored the first goal will be 67 percent. The odds that they scored the second goal will be 67 percent. And the odds that they scored the third goal will be 67 percent. So you know why teams that score first win two-thirds of the games? Because winning teams score about two-thirds of the goals. There's nothing magical about the first goal, no matter what The Stats Guy might tell you.

There are plenty of other Stats-Lite in other sports. If LeBron has a triple-double, the Cavaliers are 7-3, something like that. We've seen stats that mix cause and effec: Running teams have a better winning percentage. (Because they're ahead comfortably. So they run the ball.)

The first goal is, maybe, an indication that the team that scored it is going to be the winning team, in that they are playing better. So it's less of a priority to win and more of a sign that you have an increased likeliness that you will win.

But yeah, "record under certain condition" stats are largely useless. It used to drive me crazy hearing the Mets' record when Jose Reyes scored a run, with the idea that he scored a lot of runs and got things started as a leadoff hitter. But if you fix one player at "scores at least one run" and let the others play to their averages, you're going to win a lot of games.

Dafatone

billhoward
After Cornell got out to a 1-0 first period lead over Union, it was noted on the broadcast that if Cornell leads after one period, it has thus-and-such outcome, I think 5-1-2. You hear that a lot.

See SBNation on The Value of the First Goal

Eric T, SBNation
Teams that score the first goal win 67 percent of the time. So that first goal really sets the tone, huh? But here's the catch: teams that score the second goal win 68 percent of the time. And teams that score third win 68 percent of the time. So, uh, about that super-important first goal...

Suppose a team wins a game 4-2. They scored 67 percent of the goals. If no goal is any more important than any other, the odds that they scored the first goal will be 67 percent. The odds that they scored the second goal will be 67 percent. And the odds that they scored the third goal will be 67 percent. So you know why teams that score first win two-thirds of the games? Because winning teams score about two-thirds of the goals. There's nothing magical about the first goal, no matter what The Stats Guy might tell you.

There are plenty of other Stats-Lite in other sports. If LeBron has a triple-double, the Cavaliers are 7-3, something like that. We've seen stats that mix cause and effec: Running teams have a better winning percentage. (Because they're ahead comfortably. So they run the ball.)

The first goal is, maybe, an indication that the team that scored it is going to be the winning team, in that they are playing better. So it's less of a priority to win and more of a sign that you have an increased likeliness that you will win.

But yeah, "record under certain condition" stats are largely useless. It used to drive me crazy hearing the Mets' record when Jose Reyes scored a run, with the idea that he scored a lot of runs and got things started as a leadoff hitter. But if you fix one player at "scores at least one run" and let the others play to their averages, you're going to win a lot of games.

It's kind of like someone saying "Usually the team that scores the most points wins the game." (Supposedly John Madden said that once.)

At least now I know that in hockey (NHL) the first-to-score wins two of three times, and the winning team scores on average two-thirds of the goals in a game. No matter how the announcer tries to spin it: "They're 2-0 when Martin St. Louis comes out of the penalty box and scores on a breakaway for the game's first goal. At least when they're wearing their home jerseys excluding the retro jersey game against Detroit."

Probably does not apply to the Devils if they score first and wind up in a shootout: 2-4 in all shootouts this year, NHL record 0-12 last year and 3x43 shooting; even Brodeur wouldn't be that bad if he had a chance.

Yeah, but I do pay attention to stats that show how we protect late-leads. There was that one year where we kept blowing 3rd period leads, which up until then was an unheard of option from a Cornell team. When I see that this Cornell is 7-0-0 when leading after 2, and 3-3-3 when tied after 2, I can think to look somewhere else for "what's wrong."

Another situational stat I look at is when are our GF/GA coming from? This year, 47% (26/55) of our scoring is coming after the 2nd intermission, which suggests that either we're successfully wearing down the opposing defenses early or our offense takes too long to find a rhythm. I prefer Cornell teams that are offensively strongest in 2nd periods so we can focus on our "lockdown" strengths in the third, rather than playing Cardiac Comeback Hockey. This season, we have a -5 goal differential in each of the first two periods, and are +7 in third periods. And the interesting footnote here is that our highest scoring period is also the period we happen to shoot the least.

[www.collegehockeystats.net]