Thursday, September 22, 2005



I know there's a lot to chew on and talk about during the preseason, but I'm still a little surprised that I haven't heard any discussion at all about points: how many points will be needed to make the playoffs, win the division, the conference, etc.

I've linked this old piece of mine about a dozen times now. To recap briefly: there's no more ties in the new NHL, so every game tied at the end of regulation will be a 3-point game, and this will have a rather profound effect on the number of points associated with every level of success (or lack thereof). The average number of points per team in the NHL will be 92.5, based on the same number of games being tied (25.6%) after the end of regulation as in '03/04. Even if you think the new rules and crackdowns will create a lot more scoring, and thus fewer ties thru 60 minutes, 20% is probably the best (lowest) number you can hope for.

(In 1981/82 -- when the Oilers scored 417 goals and most goalies were hardly better than a Shooter Tutor (sorry Liut), 17.6% of games were tied.)

I bring this up again after considering the interesting pattern the Oilers have established in the 1st three preseason games (SO win, SO loss, SO win - tied thru 60 all three games).

Consider a mythical team that is tied thru 60 in every one of its 82 games. In 1982, this team would have 82 points (record of 0-0-82, 82pts). In 1992, before the OTL point was introduced, this team would also have 82 points, again assuming that they score an OT goal as often as they gave up one. (Based on the longterm consistent data that about 44% of reg. season OTs end with a goal, this team's record would be about 18-18-46, 82pts). Both these teams would roughly be on the playoff bubble, since they're mathematically average.

Now take this same team in 2002, where the standings show W-L-T-OTL. They're still winning 18 games in OT, losing 18, and tying the rest, but now their record is 18-0-46-18, 100 pts. This team is solidly in the playoffs, if maybe a hair short of "elite".

In 2005, the standings will show W-L-(OT/SO)L. This team's record will be 41-0-41, for a whopping 123pts, and they win the President's Trophy.

I'm not sure what can practically be taken from this exercise. Maybe two things:


I like your final conclusion that losing in regulation is worse than losing in OT; but, wasn't that sufficiently obvious from the fact that you get a point for an OT loss and not for a regulation loss?

In other words, I don't see the point of all your number crunching.

Thy mythical team's 100 points in 2002 does not indicate "solidly in the playoffs," and the mythical team's 123 points does not indicate that "they win the President's Trophy", unless you make the rather silly assumption that all other teams are being graded on a 1982 standard.

What is the point of making this ssumption?

Does it really matter what point total = playoff team? Does it really matter whether "0.500 ain't what it used to be"?

Am I missing something?

Besides to do this sort of analysis properly, you would need to divide the games into intra- vs inter-conference games. An OT/Shootout win against a conference rival is worth less than an OT/Shootout win against a team from the other conference. This, presumably, would play into any decisions regarding how badly you want to just survive regulation or try to put the game away in the third.

For starters, the 100-point team in 2002 is solidly in the playoffs; since the OTL point came in, I think 92 pts is the highest to ever miss out.

Also, 123 points can probably be expected to win the President's Trophy this year, even with every team earning an average of ~5-6 points more than last season.

That said, the "point" of all this is not to underline the 1-pt difference between reg.L and OT/SO L (cause yes, no math wizardry needed there); it's to underline the 1.5pt difference between losing in regulation and being tied.

Does it matter whether "0.500 ain't what it used to be"? The short answer is No, but the longer answer is, Kind Of.

Even in the years that the OTL point has existed, being above .500 has been a handy shorthand for "above average" or "on-track for the playoffs". This is way, way wrong now. If your team is 5-4-1 after 10 games this year, that doesn't sound that bad, but you're on pace for 10th or 11th place.

Guaranteed that 20-30 games into the season, the pundits will be talking about how incredibly competitive so many teams in each conference are, instead of noting that 11 or 12 teams above .500 is a statistically inevitable consequence of the method of awarding points (in fact, bookmark this comment and check it on Dec. 15).

In short, absolute point totals don't matter, but this is useful stuff to understand if you want to understand the NHL this season.

I don't think the inter/intra- conference thing makes any difference (at least, relative to last season). The point for is important no matter what; the point against is meaningless against the other conference, meaningful in your own (plus there's only 10 games against the other conference this year).

Lastly, there is virtually no circumstances under which a team this year will open it up in the 3rd to try to "put it away in regulation". 9th-place playing 8th-place in the last 5 games of the season is about it.

Reg. W = 2 pts
Reg. T = 1.5 pts
Reg. L = 0 pts

If you acknowledge that "incentives matter" then being tied in the 3rd = Operation Shutdown.

In other words, you think Hockey Fans & Pundits are so dumb, that they will not understand that looking at the actual number of points their team has relative to other teams is better than looking at the winning percentage.

You may be right; but, I honestly can't recall ever having a single NHL water-cooler conversation where winning percentage was even mentioned, nevermind used as a guide for how well the team is doing.

The fact that the NHL Standings don't typically show winning percentage may be one reason.

I think the main thing to take away from this is that there are going to be a lot of very boring third periods played this year. Particularly when playing outside your conference, there's absolutely zero incentive to take chances in order to win in regulation: the risk of losing in regulation by far outweighs the reward of winning in regulation.

Add to this the fact that a weak team with strong goalie will play to get to the shootout every night in hopes that the goalie will steal the extra point, and the new rules are a recipe for a very boring season.

Of course, the solution is simple -- grant 3 points for a regulation win. But Gary apparently didn't want to mess with the history of the game. (Which makes no sense, since, as this post clearly indicates, point totals now will bear no resemblance to point totals in the past.)

I think the main thing to take away from this is that there are going to be a lot of very boring third periods played this year.

Agreed. Matt makes this point especially well in the original post he links above.

Although I will say that I didn't really notice a drop of in third period excitement with the introduction of points for OTL (although as Matt shows, this was only worth 1.2 points vs 1.5, so the incentive to survive regulation time was less).

The other thing to note is that invariably when teams try to protect leads, their play usually deteriorates, and they end up losing it anyway. Who knows, it might be the case that both teams trying to protect a tie won't work out as well as predicted, leading to some rather interesting third periods.

"The other thing to note is that invariably when teams try to protect leads, their play usually deteriorates, and they end up losing it anyway."

That is a very un-scientific statement. Have you not read "Moneyball"? Do you have any statistics to back that thing up? Otherwise, my friend, I must declare it to be unworthy of my noting.

I'm still waiting for Scotty Bowman and a posthumous Herb Brooks to publish Moneypuck.

Post a Comment

<< Home

This page is powered by Blogger. Isn't yours?