I find the work relatively straightforward and the results reasonable -- ie, the skill results and associations with teleconnections are in line with expectations. All reviewers touch on the need for greater clarity in various areas (including further work on diction and grammer), and the authors responses are appropriate. I recommend that the author proceed with reviews as outlined and that the revised paper undergo a further round of reviews.

In addition, I find it odd that the authors do not address more comprehensively the apparent noise in skill results shown in the spatial (pixel) plots -- e.g., in Figure 9, where adjacent pixels show signals of opposite sign. Although the authors note that the differences shown are 'significant', perhaps indicating a domain-aggregate sense, I suggest the authors go further to indicate where in each season (eg, again in Figure 9) they are significant. Surely where the cell-specific differences are near zero, they cannot also be significant. It is possible that I am misinterpreting the authors' discussion, but that alone suggests the need for greater clarity and thorough effort to convey where skills that are statistically different than zero are achieved. I would hope to see this point addressed in the revision.