| From | Message |
|
Rating graph
I've recently got my rating up to an all-time high - have gone above 1900 for the first time. On my profile page, the recent history chart now extends up to 2000, because my peak rating in that period is 1918. But when I click into the all-time rating graph, it still tops out at 1900, and suggests my current rating is around 1880. Not a major issue obviously, but curious as to what's causing the difference? Is it taking an average rating over, say, the past week rather than the live rating? It seems harder to have the wrong graph than the right one!
|
|
The reason being is because the ELO Rating is not always accurate. It's reliable for sure as to how much progress you've made and where you stand against others. But, I don't go by it because Master Nathan Resika, who's coached me in OTB chess, told me not to. I'd say to go by what you think you are because that's probably what's accurate.
|
|
No, my rating is my rating. (My strength may be different, but usually not) The question is why one chart says I've been over 1900 and the other chart says I haven't
|
kingdawar 18-Aug-19, 03:39
|
My guess (after reading the graph several times) is that not enough time has passed since June for the 1900+ peak to be included. Notice each year is somehow demarcated by the year and the month June. For mine it's July.
|
|
Possible - though I don't understand how the one-year graph, which is similarly demarcated, does reflect the 1900+ peak. My own guess was that the annual graph doesn't track your rating, but some sort of weekly average rating to reduce datum points.
|
|
Curiously, I've now matched my record high of 1918, and now my main chart has jumped up to 2000. Another mystery to add to the pile!
|
|
I think kingdawar's observation is probably the answer. The coding that produces the 2 charts is probably slightly different, resulting in a difference in the timing of when new data shows up on the chart.
|
|
The overall graph often represents multiple values for each single point on the graph, so it needs to be "smoothed" by averaging the rating values. In case of cdeburca, you have 3,912 distinct rating values (finished rated games) to be shown across a graph with the width of 900 pixels. So each point on the graph represents approximately 4.3 games, or the average rating over 4.3 games. Which means you need to finish 4 games at or close to your new high rating in order for it to be shown exactly on your overall graph. The "past year" graph is more detailed as it doesn't need to include as many rating values, so each point on the graph often represents a single game (no need to average).
|
|
Cheers for that GK! As I say, not an issue for me; I was just curious.
|