The NHL Standings
Posted by
tgirsch
It really bugs me how the NHL lists its standings. Setting aside our well-documented objection to the evil OTL point, the fact that the NHL lists standings in terms of points only is just silly. As of Tuesday night, here’s how the Eastern Conference Standings look according to the NHL:
| Rk | Team | GP | Pts |
|---|---|---|---|
| 1 | * Buffalo | 52 | 72 |
| 2 | * Atlanta | 53 | 66 |
| 3 | * New Jersey | 51 | 65 |
| 4 | Ottawa | 53 | 62 |
| 5 | Montreal | 51 | 61 |
| 6 | Carolina | 53 | 58 |
| 7 | Pittsburgh | 49 | 56 |
| 8 | Tampa Bay | 52 | 56 |
| 9 | NY Rangers | 50 | 54 |
| 10 | Toronto | 51 | 54 |
| 11 | NY Islanders | 50 | 51 |
| 12 | Washington | 50 | 49 |
| 13 | Boston | 49 | 48 |
| 14 | Florida | 52 | 48 |
| 15 | Philadelphia | 50 | 29 |
Notice that there’s quite a discrepancy in number of games played. Atlanta and Carolina have played 53 games each, while Boston and Pittsburgh have only played 49. The other teams are somewhere in between. Notice, too, that Atlanta is shown as owning the #2 spot over New Jersey, despite having only one more point in two more games played. (For the NHL neophytes, a win is worth two points.) This seems wrong to me. New Jersey should be ahead of Atlanta.
In my mind, you should list the standings in terms of percentage of possible points; this gives you a better idea what the standings really are, and puts them more in line with what we’re used to seeing in baseball, football, and basketball. A sample of my suggested standings:
| Rk | Team | GP | Pts | Pct |
|---|---|---|---|---|
| 1 | * Buffalo | 52 | 72 | .692 |
| 2 | * New Jersey | 51 | 65 | .637 |
| 3 | * Atlanta | 53 | 66 | .623 |
| 4 | Montreal | 51 | 61 | .598 |
| 5 | Ottawa | 53 | 62 | .585 |
| 6 | Carolina | 53 | 58 | .547 |
| 7 | Pittsburgh | 49 | 56 | .571 |
| 8 | NY Rangers | 50 | 54 | .540 |
| 9 | Tampa Bay | 52 | 56 | .538 |
| 10 | Toronto | 51 | 54 | .529 |
| 11 | NY Islanders | 50 | 51 | .510 |
| 12 | Boston | 49 | 48 | .490 |
| 13 | Washington | 50 | 49 | .490 |
| 14 | Florida | 52 | 48 | .462 |
| 15 | Philadelphia | 50 | 29 | .290 |
Note here that New Jersey, Montreal, the New York Rangers, and Boston have all “moved up” in the standings — they’re in better shape than the initial standings would seem to indicate. This, to me, is a more accurate picture. I have no idea why the NHL doesn’t report it this way.
Is this a gripe session? I’d like to say 3 points for a regulation win, much like in soccer.
Comment 1/31/2007
Just another reason to not care about hockey, or as we refer to it, soccer on ice with sticks.
I predict OTL scoring will someday be the subject of a game theory thesis. Clearly, if teams compete in a rational way, they will collude to have all games reach overtime where the total profit (points towards playoffs) are higher than if the game is decided in regulation time. Heck, you don’t have to be John Nash to figure that out.
Comment 1/31/2007
Ted:
There’s already some game theory involved. Prior to the advent of the OTL point, games could end in ties, with each team getting one point. But too many teams would “play for the point,” not taking risks late in close games, and especially not in overtime. The idea was supposed to be that once overtime hits, you’ll play to win, since you’ll get a point anyway if you lose. Only it hasn’t worked that way.
My preference would be simply to do away with points entirely, and just do everything based on win-loss ratio. Make winning percentage be your primary record-keeper, and use fewest regulation losses as the first tie-break. But I’ve covered all that before.
Comment 1/31/2007
Wkmaier:
The problem with the three-point game is that now you’ve added another stat you need to keep track of — the overtime win (OTW), which presumably would be worth 2 points against the OTL’s 1 point. I don’t think it needs to be that complicated.
Ted:
While hockey does have some similarities to soccer, the pace and the physicality are both vastly different. I would say that hockey is like soccer on ice with sticks and with stuff actually happening.
Comment 1/31/2007
Tgirsch, your thing is kinda baseball stat, which, pardon my French, is super dorky and who gives a rat’s ass except anal purists?
Don’t take that wrong!
I guess I’m saying hockey doesn’t need to be as complex as the NHL wants it to be. So maybe I agree with you at some level?
What pisses me off is they made all those changes, and now they are talking about making the net bigger, or goalie pads smaller, or 4 on 4. Why not play the game for a few years and see what happens with no changes?
Comment 1/31/2007
Tgirsch, yes, I am aware of how hockey was scored back in the good old days. I was a Bruins fan back in the Bobby Orr days and played a bit myself in high school (before helmets but after goalie masks). I wouldn’t classify “playing for the point” as game theory (except in specific end of season scenarios where one point is as good as two). After all, assuming two closely matched teams, winning outright half the time yields the same results as always playing to a tie. And the risk of a loss relative to a tie (one less point captured) is equal to the risk of a tie relative to a win (one less point captured). If two teams colluded to play to a tie each time they met, they would not realize a net competitive advantage over the rest of the league. With OTL scoring, colluders would realize a net advantage over teams playing it straight. I doubt there is actual collusion going on though since hockey guys are clearly not bright or articulate enough to work out such a deal.
Comment 2/1/2007
Ted:
I think you oversimplify the game theory going on. It doesn’t always make sense to play for the tie, but sometimes it does. Late in the third period of a tie game, attacking (trying to score) also exposes you to defensive weakness, increasing the likelihood that the other team will score. That’s two points for them, none for you. So instead of playing to win, many (most?) coaches will simply play not to lose in this scenario. Stack the defense, don’t let them control the puck in your zone, and don’t get too aggressive.
This is even more stark when you get to overtime. Without an OTL point, once you’ve gotten to overtime, you’ve got a point at stake, which you can lose if you screw up or get too aggressive. So you wind up with ultra-conservative play, which makes for really lousy hockey. And theory aside, this is precisely what happens.
Comment 2/1/2007
Of course, if the other guy is going to lay back and not try to score then I’m not really at risk opening up some. That’s the thing about many conventional wisdoms in sports. They just don’t make sense on a rational basis.
Comment 2/1/2007
then I’m not really at risk opening up some
Baloney. If the other team is playing heavy defense, while you’re in attack mode, and you turn over a puck, there’s a very good chance of game over, because your forwards are, well, forward, and there’s nobody between the guy who stole the puck and the net. So what often happens is both teams sit back and hope the other one makes a glaring error on which to capitalize. Otherwise, they’re content with the tie. I know. I’ve seen it happen zillions of times. Instead of thinking “nothing ventured, nothing gained,” the coaches seem to think “nothing lost.”
Comment 2/1/2007
Wow, hockey has changed recently. If conventional hockey is being played, and a forward has the puck stolen, in the old days there would be two defensemen and a goalie between the puck and the net.
I’m not saying teams don’t become more conservative late in the game when the score is tied. I’m just saying it does not make sense. The score is also tied at the beginning of the game as well, so if playing offense when tied has a negative return, it has a negative return at the beginning of the game as well.
I liken the situation to playing prevent defense in football when you are less than 3 points ahead, not going for it on 4th and inches when beyond the 50 yard line, “playing for a tie” on the road but “playing for a win” at home in baseball, and switching to an acrylic broom when defending a 5-4 split if using a heavy rock on the final throw in tournament curling.
Comment 2/2/2007
Yes, but if you fall behind early, you have time to catch back up. If you fall behind late, it’s a different story. The clock is your enemy. Surely you’re not suggesting that teams should use the same strategy at 19:30 of the third as they would at 00:30 of the first.
And part of playing aggressively involves the positioning of your defensemen. If you’re playing a defensive strategy, your defensemen will rarely cross the red line. In a fairly aggressive stance, your D will be up near the opponent’s blue line, and sometimes even pinch into the play for additional offense. You’re far more open to a breakaway that way, so late in games, teams won’t do it (and, in fact, they’ll even often send in only one forechecker).
So while I agree that it doesn’t necessarily make sense for teams to become too conservative late in tie games, it most certainly is a rational transaction to become at least somewhat conservative, in particular in situations (like the NHL) where there are more options than just “win” or “lose.” It makes a world of sense from a coaching perspective: to use invented numbers, if a conservative game plan gives me a 90% chance of getting at least one point (tie or OTL), with a 5% chance of winning (two) and a 5% losing (zero points), and an aggressive game plan gives me a 30% chance of winning, but also has a 30% chance of losing (zero points), it makes perfect sense to play to guarantee at least the one point; it just makes for bad hockey for the spectators (which, contrary to popular belief, is who you’re playing the games for).
Comment 2/2/2007
Oh, and Kudos on the curling reference. I have no idea if that actually makes sense, but it’s funny.
Comment 2/2/2007
TGirsch, I agree with you. Somehow we got out of sync. I was describing strategy assuming no OTL scoring. (Remember, my original point was OTL scoring rewards collusion whereas traditional scoring does not.)
Remove the OTL consideration, and I believe the strategy of playing more conservatively as tie winds down makes no sense. The fact that giving up a goal late is more expensive is exactly offset by the fact that scoring a goal late is more valuable. As long as we are talking about a typical mid-season game between equally matched opponents and no OTL, scoring a goal to break a tie has to be a zero sum proposition when both teams are considered. The idea that it makes sense for both teams to play more defensively late implies it is not zero sum but rather there is a negative net value assigned to a goal (ie it hurts the losing team more than it helps the winning team). And that is demonstrably false. At least in terms of the standings.
I think late conservative play has a psychological basis. Just as most people of average means will not spend $9999 for a 1 in 10 chance to win $100,000 even though by doing so they increase their expected value, I think most coaches value a tie over a loss more than they value a win over a tie, even though the difference between the two is exactly the same. A sharp, rational coach can exploit this pervasive weakness and increase his team’s accumulation of points over the course of a season.
I must admit I made up the curling bit using knowledge gleaned watching about 30 minutes of curling in the last Olympics.
Comment 2/2/2007
[…] With only five or six games left (depending on your team) in the regular season, it’s time again to re-visit my previous screed on the NHL standings. […]
Pingback 3/28/2007
First, remember that the NHL used the point system to make the math calculation easier when handling ties. Then the NHL decided to give a point for an overtime loss because teams were playing overly defensively for fear of losing that ever valuable point. Of course, that made overtime more interesting for the fans now that teams actually tried to win in overtime. Later, the NHL abolished the tie by modifying the overtime rules and including a shootout if all else fails. What the NHL forgot is that they issued the “bonus” point to encourage teams to actually play for the win in overtime. Now that the tie has been eliminated, hockey should do what all other sports do which is Win/Loss percentage. Afterall, if your favorite baseball team loses in 27 innings, there is no extra credit because of how long it took you to lose. Your favorite basketball team loses in 4 overtimes, same deal, no extra credit. Why does the NHL feel that their teams deserve extra credit for losing in overtime? Montreal and Colorado (neither team of which I’m a fan) were both eliminated from the playoffs because of the NHL’s rules and not because of their performance on the ice. It is time to change the NHL’s rules to one that is fair for all teams and not be the only sport to give credit depending on how you lose.
Comment 4/9/2007
[…] That’s right, the NHL regular season has ended, and our loyal readers know what that means: it’s time for me (and probably Kevin) to bitch about the OTL point again. For prior OTL hatred, see here; here; and especially here and here and here. […]
Pingback 4/8/2008
I like the OTL system, because it is like a tie. And somehow American sports culture has decided to outlaw the tie. There used to be lots of famous ties like the 10-10 tie between Michigan State and Notredame in football. Or Montreal Russia New Years eve tie in hockey.
Most of the world accepts regular season ties in Soccer. But America has made it impossible in it’s mainstream sports to have a tie. Hockey is the one holdout. I was initially agains hockey getting of the tie for the OT format they chose. But I find it is one of the more exciting formats now.
Comment 6/10/2008
But it’s not like a tie, though, because in a tie, both teams get equal credit; but in this system, the team that wins in overtime isn’t punished for taking longer to win, while the team that loses is rewarded for taking longer to lose. In my mind, you shouldn’t have one without the other.
By the way, you can still have a tie in American football — it just doesn’t happen very often. We came within 3 seconds of one this last year, however.
Comment 6/25/2008