When there are long breaks in between Leaf games, such as what us Leaf fans are currently experiencing right now (the Leafs play their next game on Friday, November 26th, and their last game was Monday, November 22nd), Toronto sports writers often times struggle to come up with good, captivating content regarding the Leafs such that readers of the daily newspaper have some Leaf-related content to keep them coming back to the sports section. All the major newspapers in Toronto do this (Toronto Star, Toronto Sun, Globe, etc). Sometimes, the content which is dug up and regurgitated on non-game days is actually quite good, but other times it’s a bit nonsensical.

In today’s Toronto Star (November 25th), Kevin McGran has written an article entitled, “Statistics Show Leafs’ Playoff Chances Slim“. In this article, McGran compiles statistics dating back to 1993 that analyze the probability of a team making the playoffs based on their current point total and position in standings (ie. playoff position or not). Based on Kevin’s findings, it would appear as though the Leafs (who currently have 19 points in 20 games, and are not in a playoff position) stand a 12% chance of making the playoffs. According to Kevin, only 2 of 17 teams who had this many points at this point in the year have ever gone on to make the playoffs. He also states that 84% of teams that had 17 or fewer points at the U.S. Thanksgiving mark also failed to make the playoffs. What’s more astonishing, according to Kevin, is that having just 2 more points (ie. one extra win) pushes you into the playoffs 59% of the time.

While I respect Kevin and have enjoyed reading his articles in the past, I really don’t believe that the numbers he states, while based on real data, are representative of true “probabilities”. For instance, Kevin presents a “probability” table which indicates that the Leafs stand a 12% chance to make the playoffs based on their current point total and history. However, the same table also states that teams with 18 points at this time of year stand a 30% chance of making the playoffs. If you ask me, that’s a bit of a head scratcher. How can a having one less point in the standings, at this time of year, give you a more than 100% boost in your chances of making the playoffs? The same table Kevin presents also shows that having 16 points at this time of year (3 less than the Leafs current total) brings you to a 36% chance of making the playoffs. At 13 points (6 less than the Leafs current total) brings you to 29%, and finally, having 10 points (9 less than the Leafs current total) brings you to 33% chance of making the playoffs.

Reviewing Kevin’s table and his conclusions that the leafs “probability” of making the playoffs are 12%, based on their current point totals and history, makes absolutely no sense to me when the same table indicates having less point totals yield a higher chance of making the playoffs. Kevin’s table really shouldn’t be called “probability”, but rather “historical results” or something along those lines. A probability indicates the likelihood of an event or result occurring, and while Kevin’s numbers are backed up by historical data, they are completely random results. In other words, there is no scientific or mathematical basis which points to the fact that if you have 18 points in the standings now, you stand a better chance to make the playoffs then another team (the Leafs) who have 19 points. Sure, these are the numbers that have occurred in history, but it is completely random. I wouldn’t put much stock in Kevin’s probability table whatsoever.

Secondly, Kevin’s table also indicates that if you have 21 points at this time of year (remember, that’s 1 more win than the Leafs currently have), your chances of making the playoffs skyrocket from 12% to 59%. Again, historical data shows this, but it’s not a probability and has no mathematical basis behind it.

What does all of this mean? Well, for one thing, Hockey is a completely random thing. There are so many factors that come into play regarding a team’s final placing in the standings. Key players (your star goalie) can go down in injury, other players can become incredibly hot (or cold), a coaching change mid-season, a big trade, etc. All of these factors, and many many more, can turn the fortunes of any team. People always search for clear indicators, numbers, which can be used to adequately predict the outcomes of Sports. Heck, I do it all the time. At the end of the day, there is no true means of assessing the outcome of a sports match-up, or a team’s final placing in the standings and whether or not they make the playoffs.

Now, do I make this argument because I’m furious at Kevin that his statistics show the Leafs will likely not make the playoffs? Not at all, and in fact I still don’t believe the Leafs are in the top 8 of teams in the East, but neither do I believe that the method by which Kevin has followed to convince Leaf fans that the Leafs stand little chance to make the playoffs, and neither should you.

Heh, great breakdown of this article. I, too, like Kevin McGran but maybe his editors need to be more familiar with statistical models before this stuff goes to publishing.

Hi Karina,

I agree with you 100%. However, I suspect that the Toronto Star would let this one slide given the fact that it would cause fans to “panic”. Panic and fear brings more readers. Unfortunately, the source material for this article is very weak, yet I suspect the majority of readers will buy into it.

Excellent points. You nailed it when you point out that McGran’s historical numbers do not equate to probability.

To me, the biggest problem is that the teams have all played different # of games, so using points alone — even if current win/loss trends remain static, which they won’t — is not a proper measure of their playoff potential.

A more useful analysis might look like this:

1) select a milestone # of games — say 20 games, to approx represent a quarter-season

2) rank teams by points-per-game-played, for the current season vs. past seasons

3) using this ranking, plot a chart showing % of teams who made the playoffs … that way outliers on the curve (such as McGran’s supposed 12% success rate for 19 points @ Thanksgiving) are not misleading

What do you say Craig, any interest in re-interpreting and doing a plot like this?

Keep up the fantabulous blog.

Cheers,

Adam

Oops, what I meant to say was:

1A) select a milestone # of games — say 20 games, to approx represent a quarter-season, rank accordingly

– OR –

1B) for a given date (e.g. Thanksgiving) rank teams by points-per-game-played

2) using this ranking, plot a chart showing % of teams who made the playoffs in the past vs. the current season

This (McGrans’ article) is what happens when there’s nothing to write about; writers start scrambling for anything they can put together a few hundred words about, and it’s usually garbage.

Just as well to wait it out. Any time I’ve ever tried to force an article I usually end up reading it over and hitting the ‘delete’ button. I don’t even bother with it anymore.

This, however, is a great article.. so keep up the good work!

Hi Ryan,

You are bang on, and I agree. You see all sorts of poor articles surfacing whenever a writer is forced to publish some content when there’s nothing really new to report about. Whenever that occurs, the quality suffers.

Many thanks for the kind comments!

Exactly.

When I first started my blog sometimes I’d force content right after an article that drew a lot of traffic to try to “ride the wave” I guess. Turned out some brutal content for sure.

Funny that you mentioned this today though. I was thinking about it this morning and mentioned it on my blog how there was basically nothing interesting to weigh in on surrounding the Leafs in the past couple days.