Hi everyone -
The initial roll out of the "Top Solvers" lists was a bit screwy - thanks to LLapp and others who pointed out the bugs. I've spent the last day recompiling all the old solve data and I think I've got it sorted now. The final script is currently running which means puzzles may have conflicts "All Time Record" and "Top Solvers" data showing for the next 1-2 hours while it is finishing up processing all our quotes. By around noon eastern today I expect we should be all caught up and the stats should be accurate.
Some people were asking why 2014 was poorly represented in the data... its because we've had a few glitches over the years where databases have crashed and backups got corrupted. The new "Top Solvers" lists compiles about nine different SQL backups I've been able to cobble together from the last 12 years. It has *most* of the solve data from the site but some unfortunately is either lost for good or in a completely unusable format due to data corruption- 2014 being a good example. The only data we have from that year (mostly) are new all-time records that happened to be set in 2014. So you'll see a few scores in there from that year, but not very many.
We standardized on a much better data schema in 2015 so all data from then on is complete.
Thanks!
Stephen
The initial roll out of the "Top Solvers" lists was a bit screwy - thanks to LLapp and others who pointed out the bugs. I've spent the last day recompiling all the old solve data and I think I've got it sorted now. The final script is currently running which means puzzles may have conflicts "All Time Record" and "Top Solvers" data showing for the next 1-2 hours while it is finishing up processing all our quotes. By around noon eastern today I expect we should be all caught up and the stats should be accurate.
Some people were asking why 2014 was poorly represented in the data... its because we've had a few glitches over the years where databases have crashed and backups got corrupted. The new "Top Solvers" lists compiles about nine different SQL backups I've been able to cobble together from the last 12 years. It has *most* of the solve data from the site but some unfortunately is either lost for good or in a completely unusable format due to data corruption- 2014 being a good example. The only data we have from that year (mostly) are new all-time records that happened to be set in 2014. So you'll see a few scores in there from that year, but not very many.
We standardized on a much better data schema in 2015 so all data from then on is complete.
Thanks!
Stephen
Comment