## How Many Points to Make the Playoffs? A Statistical Analysis

Like all of us, seeing the Devils drop 7 in a row and leaving lots of points on the table is scary. How can the team that was in first place and a year prior was 2 games shy of a Stanley Cup not make the playoffs? I put the numbers to work just to see how many points we need to be the 8th team in the playoffs. If you don't like statistics and just want to know the answer go to the end.

First I compiled a historical data set of all NHL seasons since the shootout was introduced. The problem with doing this is that the two conferences play each other, so if one conferences plays better against the other (which happened every year between 05-06 and 10-11 where the west had a wide advantage). Last season, interestingly, the total points in the western conference only outnumbered those of the east by 4, or 2 wins, while both conferences had an equal number of 3 point games at 150 each. Side note: this means that last year at least the West was not better than the East, as "experts" would have you believe. But it takes more points to get in the playoffs in the west? Well more points against the east is one way the west has achieved that, but another is the variability of the west in the standings. Wins happen in a vacuum, one team wins and the other loses. The variance of teams in the west in all years but 08-09 was greater in the west than the east. In real teams that means the teams at the top win more than the top teams of the east, but the teams at the bottom also lose more than the bottom teams of the east. My point is that the amount of points needed to get in the playoffs is not what we should use to compare leagues, but only their records against each other. But there are problems with that too that I won't get into.

Back to the issue at hand, to overcome the problem of the leagues playing each other, I averaged the amount of points needed to make the playoffs between the conferences for each season. This is not ideal but it is all we have to work with. I cannot take just the points needed to make the east, because there is western interference in that number. This number for points to make the playoffs is my dependent variable, for which I look at the relationships of all other numbers for a season too, because that is the statistic that answers my question for this season. I hypothesized that the amount of 3 point games, the average points earned in a conference, the variance of points in a conference measured by standard deviation, and the median points in the conference would predict how many points are needed to make the playoffs.

There have been 75 three point games in the west and 60 in the east. This is the first time ever in the shootout era there have been more in the west, let alone a lot more. Because we can expect it will take more points in the west, since more points have been available to earn. There is also a greater variance in standings points in the west. Easily explained by there being two top dominant teams in the west that rarely lose, resulting in bottom teams in the west that are worst than the bottom teams in the east. Because of this we can be sure it will take more points in the west, but since there is no interconference play that does not mean the west is better. It could mean that the east is more evenly matched, or that the bottom teams in the west are much worse (Colorado, Calgary) than the sun belt teams in the east who make up the bottom.

The average points number as a result of 3 point games is our measure of central tendency. Instead of .500 being a mark to judge winning teams, the average in the NHL is better. Before the shootout with ties, the 8th seed was usually the last team in the conference over .500, a winning record. but now the 9th and 10th seeds can be winning teams too and miss the playoffs. It really takes a winning percentage or points percentage of 56-57% to make the playoffs. The average however is likely to be around 53%.

The median is also a measure of central tendency in the standings, and it is the best. in a 15 team conference the median is team number 8. Obviously that is also the threshold to make the playoffs. So whatever value the median has, is the result that we are looking for. We could simply look at the points per game percentages of all the teams, multiply them by the 96 points available to them in 48 games this season, and take the median of the group as the number of points needed to make the playoffs. In the east this number is 51.7 and the west is 52.9. These numbers are unreliable though, a team can always "pick up the pace" so to speak. But these are good numbers to compare a statistical prediction too.

To arrive at my answer I put my calculated numbers from each season into a regression equation, with points needed to make as the dependent variable. The computer tells me that the only variables worth using are average points, standard deviation of the standings, and the median points. The prediction equation I get can predict with 94.1% accuracy, which considering there are only 7 seasons to go off of is phenomenal.

There is another way to find an equation if you are unsure of what is best like I was called stepwise regression. This is something you use in absence of a theoretical framework, very appropriate for this situation, when it is usually not. The prediction equation that it gave me included only one variable, median points. As you remember that this is the actual number we are looking for. The median in this season of a closed conference is the amount of points needed to make the playoffs. But in the past when you take all 30 teams and calculate a median, you don't get the exact result, but you do get a 99% accurate equation. This could be inaccurate for the same reason a simple prediction based on points per game is inaccurate. One statistic is more subject to change in the last 3 weeks of the season than 3 statistics are. I will use both equations though to make a prediction.

The Results You Are Looking For

The equations for the eastern conference yields that it will take 52.5 or 52.9 points to make the playoffs. The current pace (unreliable statistic) says that it will take 51.6 points. The two equations use history as their guide, so I would go with them that it will take 53 points to make the playoffs. Either way however, 52 or 53 points should be the number. That means the Devils have to go at least 6-3 or 5-2-2 or 4-1-4 but really need to go 6-2-1, 5-1-2, or 4-0-5 to make the playoffs. And at this point wins on Toronto and Ottawa would be nice and the 2 games against the Rangers are must wins.

There you go, that's what the Devils have to accomplish. With 80% of the games played in the eastern conference I can be 80% sure of this number.

If you are curious it will take 54.7 or 53.7 points to make in the west where there is a current pace of 52.9. The two equations are more different that the difference in the east because there is more variability in the western conference standings.

All FanPosts and FanShots are the respective work of the author and not representative of the writers or other users of In Lou We Trust.

## Trending Discussions

forgot?

We'll email you a reset link.

Try another email?

### Almost done,

By becoming a registered user, you are also agreeing to our Terms and confirming that you have read our Privacy Policy.

### Join In Lou We Trust

You must be a member of In Lou We Trust to participate.

We have our own Community Guidelines at In Lou We Trust. You should read them.

### Join In Lou We Trust

You must be a member of In Lou We Trust to participate.

We have our own Community Guidelines at In Lou We Trust. You should read them.