The Statistical Relationship Between Championships and Hall-of-Famers

[Corco 1,281]

Is There a Mathematic Correlation Between Hall of Famers Playing for Your Franchise and Your Franchise Winning Championships?

 

Before I dive into the results of the above question, I want to explain what I mean by "mathematically significant". In statistics, you can measure the relationship between two variables with a complex formula that Excel can [thankfully] calculate so long as you have the data. The resultant of the formula is a number between 0 and 1, with the scale below showing the strength of the relationship:

I plugged the data into Excel to see how strong the relationship was between having a Hall-of-Fame player playing for a franchise and said franchise having historical success winning championships. I only included franchises that were around for a decent amount of time in order to remove outliers (sorry Malmo, Moscow, DC, and Prague). These are my results:

 

Franchise	Championships	HOF Players	Playoff Appearances	Seasons
Calgary	9	31	41	68
New York/Hamilton	7	35	38	68
Seattle	8	33	36	68
Toronto	9	36	39	68
Davos/Amstel/Avangard	10	34	39	68
Helsinki	11	36	39	68
Riga/Stockholm	8	31	44	68
Vasteras/Madrid/Stockholm	2	28	24	57
Vancouver/Quebec City	3	17	25	37
Cologne	1	10	15	27
Correlation between Championships and HOF Players	0.84
Correlation between Playoff Appearances and HoF Players	0.87

 

As you can see by the correlation values above, there appears to be a strong relationship between the amount of postseason success a franchise has and the amount of Hall-of-Famers who have played for said franchise. Despite the numbers backing up my statement, I believe these findings are a bit flawed, for a number of reasons.

 

1. This test is usually for larger data sets; the VHL had only 8 teams for 30 seasons, meaning that there was a theoretical (purely mathematical, with no external factors affecting the probability) 12.5% chance a HoF player would play for your franchise up until S31, a decently high number in terms of probability. And when I ran the numbers for those players who finished their career before the S31 expansion, the distribution was pretty even, see the chart below:
   
   Continuing on my point above, the theoretical probability that you will win a championship in the VHL is also 12.5% (using the data from just S1-S30), which is massively larger than the NHL for example, where your theoretical probability of winning a title is a mere 3.2% (with the current 31 teams; the chances will lessen to 3.1% when Seattle joins the league). Having such a high chance to win a title in any given season, compounded with that same high chance for a HoFer to play for that same franchise, will increase the strength of the relationship.
    
2. This is perhaps the biggest correlating factor between my two variables; Hall of Fame players are good. Good players = more victories; more victories = more championships. You would expect there to be a strong correlation between the two, in fact I would have been very surprised if the correlation wasn't strong. 

 

 

For week of 11/18/2019-11/24/2019

[Victor 11,043]

This is hawt.

 

A further point is of course playing for a championship-winning team helps HOF credentials. Not just for the Continental Cup on the resume (we have a fair few cupless HOFers), but just because those will generally be the teams with better chemistry and therefore better individual performances. So it's almost a case of which takes precedence?

[CowboyinAmerica 2,918]

I always appreciate articles that start with a mandatory explanation of r/r-squared. (Actually not a joke.)

[zepheter 590]

Hey if I'm on a good team my chances will be increased. This is great news.

[Motzaburger 1,590]

That sig tho