You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
The colour gradient used for the cells is currently ineffective in differentiating between the two extremes, which, though uncommon, do occur with poor sources, like the winrates for nations+BR combinations with very low flyout numbers. The gradient should at least be constrained (for winrates, at a minimum) to between 20% and 80%, as any cell outside of these can be safely assumed to not represent reality outside of collected data - that is, there is nothing meaningful communicated by a visual difference between a cell with an 80% winrate and a 100% winrate, as both are likely coming from incomplete data.
The graph shown when a cell is selected currently has its x axis going all the way back to December of 2019, but only displays data back to April or May of last year. Why? Almost three quarters of the graph are completely unused, and the x axis' is crowded with the 3-month increments to the point that they're almost illegible.
Finally, I highly recommend that at some point additional documentation is provided on-site to assist with interpretation of the statistics. I often post excerpts to discord servers in the community to shed light on balance and player skill across nations and BRs, but I find I need to do a lot of work explaining inconsistencies in the data, and ultimately making conjecture on the inner workings of the site, because so little is explained in explicit language. As one of the main community sources for average winrates and player populations, this site is invaluable for providing some modicum of transparency against the black wall that is Gaijin's balance decisions. In pursuit of that goal, providing tools to assist the average player in reading, comprehending, and interpreting this data is the most valuable improvement you can make.
Sorry for the wall of text - these are just suggestions, of course. I get worked up about data communication in general, but especially in this context, since the balance in WT has been so temperamental lately.
The text was updated successfully, but these errors were encountered:
The colour gradient used for the cells is currently ineffective in differentiating between the two extremes, which, though uncommon, do occur with poor sources, like the winrates for nations+BR combinations with very low flyout numbers. The gradient should at least be constrained (for winrates, at a minimum) to between 20% and 80%, as any cell outside of these can be safely assumed to not represent reality outside of collected data - that is, there is nothing meaningful communicated by a visual difference between a cell with an 80% winrate and a 100% winrate, as both are likely coming from incomplete data.
The graph shown when a cell is selected currently has its x axis going all the way back to December of 2019, but only displays data back to April or May of last year. Why? Almost three quarters of the graph are completely unused, and the x axis' is crowded with the 3-month increments to the point that they're almost illegible.
Finally, I highly recommend that at some point additional documentation is provided on-site to assist with interpretation of the statistics. I often post excerpts to discord servers in the community to shed light on balance and player skill across nations and BRs, but I find I need to do a lot of work explaining inconsistencies in the data, and ultimately making conjecture on the inner workings of the site, because so little is explained in explicit language. As one of the main community sources for average winrates and player populations, this site is invaluable for providing some modicum of transparency against the black wall that is Gaijin's balance decisions. In pursuit of that goal, providing tools to assist the average player in reading, comprehending, and interpreting this data is the most valuable improvement you can make.
Sorry for the wall of text - these are just suggestions, of course. I get worked up about data communication in general, but especially in this context, since the balance in WT has been so temperamental lately.
The text was updated successfully, but these errors were encountered: