Lila: More trophies for marathons

[nojoking]

The size of the participation in marathons has grown 5x or 10x while the number of trophies awarded remains the same. I think interest in marathons would increase if we gave trophies to the top 200 or 300.

[Whitedancingrockstar]

The most recent marathon saw less than 0.4% of players get a trophy compared to ca 3% in the first marathons from 5 years ago. Following the trend of top 10, 50 and 100, I suggest that top 500 and top 1000 also get a trophy starting next time.

[tmmlaarhoven]

To elaborate with some data, below are some graphs comparing points required per ranking for different marathons held over time. The below graphs only compare marathons with the exact same time controls, so we are really comparing apples and apples. These graphs are only for 3+0 and 3+2 as those time controls were used most often and have the most data.

The fifth figure, comparing marathons for 3+2 and looking at ratios between rankings is perhaps the clearest -- on average if you scored N'th place in winter15, that would have gotten you around (3*N)'th place in winter19, autumn19, and winter18. So ignoring the fact that the average ratings in the top of the rankings have likely gone up considerably over time, just looking purely at points/ranking, getting top 100 in 2015 was roughly equally hard as top 300 now.

[tmmlaarhoven]

One more graph below, comparing (smoothed out) blitz ratings per ranking for the (autumn) marathons which were all blitz rated. Roughly speaking with 2000 blitz you could get top 50-100 in 2015, while now your expected ranking might be around 2000. And while 2100 was roughly good enough to make top 100 in autumn 2016-17-18, in 2020 that might get you around 1000th place.

(The smoothing might slightly distort the picture; a point at {x, y} means that at ranking x, averaging the ratings of players ranked x, x+1, x+2, ..., 1.2*x, the average rating is y.)