The importance of Thanksgiving or the Kings MUST be in playoff position by Nov 23
I recently read in one of John Buccigross' mail bags that the standings don't change very much after Thanksgiving. My knee jerk reaction to this statement was that it was patently false. I thought I could remember the standings being quite fluid throughout the year. So I decided to crunch the numbers and see. I compiled the standings of every team on Nov 23rd and the standings of every team as of the end of the season for the past seven seasons. (I had to use ESPN's schedule per team, per year and manually enter the numbers and they only went back seven years. If anyone knows a place where I can find more information on more seasons I would be quite grateful.)
I picked Nov 23rd for a couple of reasons. The first and foremost being that it is always right around Thanksgiving. I didn't want to pick the actual Thanksgiving day because it changes each year. And secondly I picked it because at that point almost exactly one quarter of the NHL season is over. Teams have played enough games that small sample size errors begin to go away. I also based all my data on the position of a team in the rankings. Not by team. I wanted to find out how #1 seeds did not how the Kings or any other team performed. Because the tie break rules change from year to year I computed my standings with the following tie break system: 1, points; 2, fewest games played; 3, wins. I also want to stress that I ignored Division standings. This is important because I wanted to gauge where teams were relative to the rest of their conference. I didn't want a team that, as of Nov 23rd, was ninth in their conference and second in their division ::cough:: Southeast ::cough:: to seem like they jumped six spots because of the end of their season they caught their division leader. It would skew the data.
I went looking for several pieces of information. I wanted to find how many points a team gained in the first quarter of the season compared to how many points they would have as of the end of the season and on average how the appeared based on seeding. I wanted to find how many places, on average, a team would change based on the end of year standings. I wanted to find how often a team made the playoffs based on the amount of points they were ahead or behind of the the eighth spot as of Nov 23rd.
The short version is that Buccigross was right. If you are in the playoffs as of Nov 23 you have 75% chance of making the playoffs. If you're not in the playoffs as of Nov 23 you have a 30% chance of making the playoffs. Obviously the higher you are in the standings the higher the chances of making the playoffs. And even the eight seed had a nearly a two out of three chance. Once you drop out of the top eight only the nine seed has a better than fifty/fifty chance of making it.
I thought the most important data was how far ahead, in terms of points, a team was determined how well they did by the end of the season. Teams that were a mere five points ahead of the nine seed by Nov 23 make the playoffs over 80% of the time. Teams that were one to four points ahead of the nine seed make the playoffs just more 67% of the time. If a team is either tied with the nine seed or just four points behind the nine seed they have just under a 40% chance of making the playoffs. If a team is five points behind the nine seed or more they make the playoffs just 10% of the time.
As for the change in position of the standings there is a VERY strong regression to the mean every single year. The biggest differences were at the one a fifteen seeds and they got smaller and smaller as they approached the eight seed. No seed moved more than two spots on average except for the fifteen and two seeds which moved just over three spots. This makes sense for the fifteen seed as you can't go down at all. The knee jerk reaction would be the same for the one seed; however, the one seed can be far enough ahead that it doesn't lose much standings wise while the two seed doesn't have the same cushion. If no seed moves much more than two spots over the remaining three fourths of the season there can't be much change in the standings overall. I will say that the biggest single season move up was by Pittsburgh in the 07-08 season. Going from 13th to 2nd. They went an astonishing 38-16-2 over the remainder of the season. The biggest single season move down was by Carolina in the 02-03 season. They went from fourth to dead last in the east. Finishing a miserable 13-38-7-3. At the end of the season they had 61 points ... they earned 25 points in their first 21 games. Ouch.
Lastly, the amount of points a team earned by Nov 23 strongly correlated to the amount of points a team would have by the end of the season. Once again though I must say the regression to the mean had a very powerful affect on these as well. In this case one seeds did the worst while the fifteen seeds did the best. The numbers were almost perfectly inverse with the one exception of nine seeds. They outperformed 10s and 11s an did just a hair worse than the 12s. I would imagine this can be attributed to the urge to make the playoffs when they are so tantalizingly close. Interesting though that the nine seeds still don't make the playoffs more often than the eights do.
So what does this mean for the Kings? Well quite simply the Kings are in a great spot right now. They are four points ahead of ninth place Calgary. If they can get that up to five points over the next nine games they are virtual lock for the playoffs. Make sure that they stay ahead of whoever is ninth. If they fall behind they are in serious trouble. If they get it up over 10 points they should petition to not have to play the remainder of the season as no team has yet to be that far ahead as of Nov 23 and not made the playoffs.
One final note, I had to manually enter all the date. And although I did double check everything there is still a chance I could be off slightly.
I'm sure there's much more information that can be gleaned from this and if anybody wants to see the data just send me a PM and I will e-mail you the spreadsheet.
Shouldn't the 75% + 30% = 100% in the above statement?
You have a little free time on your hands, eh? I'll take the executive summary please
It doesn't quite work like that. There are eight teams that make the playoffs every year and seven teams that don't. Thus the differences in percentages. For example let's say one team that was out of the playoffs as of Nov 23rd made the playoffs at the end of the season. Seven teams out of eight made the playoffs which is 87.5%. One team out of seven made the playoffs that weren't in it as of Nov 23 which is 14.3%.