Sure. Basically, it evaluates the relationship between the two metrics listed, at the given salary cap (or overall with "All"). The value is expressed between 0 and 1.
A number closer to 1 indicates that a stronger relationship (you can think of this as "as Pre-Rank goes down, Tournament Rank is likely to go down as well"), while a number closer to 0 indicates a weaker (or barely existent) relationship. While there's no set cutoffs, a value under .3 is generally considered weak, while a value over .7 is generally seen as strongly related.
From this, we can get some interesting insights. So, for example, the league where Pre-Rank is most closely related to wins in that league was the 255M league, and the league where it was least related to wins was the 100M league. Another example is that we can determine that wins in the 110M league had the strongest relationship with overall tournament rank, while the weakest was again the 100M league.
Overall, we see a trend where Pre-Rank did have strong predictive power on success in the tournament, but had the least influence on the 100M league, where success was less related to pre-tournament indicators. This makes sense since it is also the league that saw the least variance (in terms of win totals) between the best and worst teams, but is still interesting in its own right since, arguably, the 110M league was more research-intensive.
10/12/2018 11:28 PM (edited)