The changing relationship between goals and winning in Gary Bettman's NHL

The greater parity is due to the salary cap AND the bonus points awarded for reaching OT. Which is the greater factor is an open question.

if i am not wrong that is GF/GA

1) Thanks Dale. I noticed the deficiency and had just edited to add a definition, but Sum has got it: Goal Ratio is GF:GA and Goal Differential is GF – GA. Usually but not quite always the best team at one is the best at the other. But to illustrate I’ll use one of the seasons where the leaders in the two categories were different:

1985-86 Edmonton Oilers: 426 GF, 310 GA
GR = 426/310 = 1.374
GD = 426 – 310 = +116

1985-86 Philadelphia Flyers: 335 GF, 241 GA
GR = 335/241 = 1.390
GD = 335 – 241 = +94

The Oilers were more prolific and outscored by a wider gross margin, but because their games were higher scoring each individual goal had slightly less value which is where Goal Ratio is a useful alternate measuring stick.

2) The Blackhawks had a goal ratio of 262/203 = 1.291 in 2009-10. More than all of that margin was on the shots clock, where Hawks outshot their opponents by a massive 2798 to 2058, for a shots ratio of 1.360. Their shooting percentage was actually slightly worse than their opponents at 9.36% vs. 9.86%, for a Sh% ratio was 0.949. So they would be plotted at X = 1.360, Y = 0.949, out towards the right edge a little below the centre line. They’re not quite as extreme an outshooter as the 2007-08 Red Wings, who are the square out near the right edge of the diagram, but they would cozy up next to the first large diamond a little further in, which represents the 2003 New Jersey Devils (X = 1.348, Y = 0.965)

3) It’s true (some) teams play different in the playoffs than the regular season. That was why I was unsatisfied in Part One that I was judging Stanley Cup teams based on their regular season shots and percentages. Hell, they don’t even play by the same rules in the regular season! I plan to address this in a future instalment(s) of my Shot Qual v. Quan series, by specifically examining playoff shots data. Apples v. apples and all that.

Thanks TartanBill. Back in the day I used to read the Bill James Baseball Abstract from cover to cover two or three times a season, learned a lot about ways to think about numbers. James (used to?) refer to the bounceback from teams that serious over/under performed their Pythagorean expectation, as the Johnson Effect.

But it sure in hell doesn’t explain the Ottawa Senators!

Thanks, Scott, that is an excellent idea which has given me an idea of my own as to how to address it. As you know I’m more about interpreting past results than I am into predicting future performance, but in this case we could use known regular season performance to “predict” immediately subsequent known playoff performance. Or not. It’ll be interesting to see if there are Any strong trends.

Which “SP” are you referring to, btw? Sh%, Sv%, or some combination of the two (PDO #, C&B #)? It’ll be interesting to see if your assumptions hold. Of course GD should be a higher order indicator since (to oversimplify) it’s measured in goals, not shots. And it seems to me the team with more goals wins a lot more games than the team with more shots. :D

I’m pretty sure that Bruce’s point is that they aren’t as bad as their record implies and that records in general are becoming a less useful measure of which teams are actually good at playing hockey.

There’s little correlation as a general rule, esp. when it comes to shootouts. The ‘07-08 Oiler still hold some sort of record with their 15 wins in 19 shootouts in ’07-08, which was the only reason the team even got within sniffing distance of the postseason, cuz they weren’t very good at all at the 5-on-5 game. Whereas Ottawa was arguably the best club in the league in the years around the lockout, and they went 2-6, 2-6 in the first two years of the shootout. So they cost themselves a few standings points right there with results that I didn’t even consider! I can’t get past 1 “goal” on 0 shots, the data is corrupted right there, so for my shots project the shootout goals had to go. But their effect is right there in the standings, which I am basically arguing have themselves been corrupted.

Btw, I followed your lead and looked up Ottawa’s record in one-goal games in 2003-04, the year they finished sixth despite leading the league in both GD and GR. Sure enough, they posted an abysmal 9-17 mark, including losing their first TEN one-goal games of the season. Their season started with these scores (OTT first): 5-2, 2-3, 3-4, 3-0, 4-1, 5-1, 6-2, 2-3 …

Now whether that is just random dice coming up snake eyes or whether that’s some sort of collective character flaw inherent to that Ottawa Senators team I’ll leave for you to decide. ;) Kind of interesting that it happened to the same team more than once.