Basketball is a sport that has inspired frenzied debate, but there has almost always been a baseline for determining quality, and that is statistics. One of the most complete resources with regard to sports statistics is the sports reference family of sites. The most easily accessible features of the site when looking at stats from a specific player are the per game stats and total stats. A conundrum might begin to face basketball fans because of two simple issues. The first is that per game stats do not show the full story because of the fact that humans get tired and are not able to play a full game of basketball. The other is that career stats are not very easy to compare, because players’ careers are of differing lengths. Even if seasons are looked at, the first issue appears, namely that stats per game (per season stats are essentially points per 82 (or however many) games) do not show minutes in the right way. To use an example from recent basketball, of course LeBron James had the most points in the 2017-2018 season; he also had the most minutes. The question then becomes one of what should be used instead. The simple answer is minutes per and per minute stats. While naysayers might claim that these are, respectively, an inverted reduction and a reduction of the already displayed per 36 minutes stats. However, there is a crucial difference. Per minute stats are a lot more applicable easily for individual games. Again going to the 2017-2018 season, only 12 players played for more than 36 minutes in a game. In order to look at the expected performance of a player based on this stat, additional work is needed to find that and how it stacks up compared to the actual performance. With per minute stats, all that is needed is simple multiplication. As for the inversion, it is a lot easier to visualize simple stats such as how it takes Muggsy Bogues 652 minutes (or 10 hours and 52 minutes (or 13 and a half full games)) to get a block like that than saying that it takes him 18 intervals of 36 minutes to get a block. The latter is simply pointless and difficult to read. However, this can only ever show the results in one category. It seems impossible when instances like the aforementioned Muggsy Bogues exist to create a stat that shows all. Not every stat exists on an equal playing field. It’s significantly easier to get points than to get blocks. In order the counteract it, I created a simple equation to determine how good a player is based on the five stats that lie at the heart of basketball: points, assists, rebounds, steals, and blocks. I first created a data set by taking the top 100 players of each of these stats, and removing the ones who were duplicated or whose careers started before the 1973-1974 season (when steals and blocks began to be recorded). I then added a few players, based mostly on nicknames and notable players not included in the totals lists (e.g. Brad Daugherty). I then weighted the numbers based on how much they were present in the total number, using points, by far the largest number, as a baseline. These weights were then multiplied by a second, arbitrary weight based on usefulness (points and assists were worth 1, rebounds and steals were .75, and blocks were .5). I then added these numbers together and divided them by total minutes played to determine the greatest (or at least most well-rounded players) in my dataset. This is better than efficiency rating because it values defensive statistics and it actually deals with minutes instead of games. This also disregards pace, because pace does not work with minute stats like these. Possessions are taken away in favor of the almighty minute. They are already factored in. What results is the list displayed on the opposite panel, a mosaic of the top 60 players of all time. Some are old, many are new and some are unbelievable, but they are all proof of the power of statistics.
Create your own unique website with customizable templates.