While not totally related, the order of features drawn on the web map affects how they are displaying when using a time slider- and this is misleading when trying to show changes by attributes. For example, my web map has a feature layer w/ 4 features per location since that is how many records were in my related table-- my published feature class essentially quadruples in size as my map service since each related record creates a separate feature in the service. I put this layer into a web map and enable the time slider-- and the display of features is different than what is shown in the time slider in ArcMap. For example, a red point for a graduated symbol = 100% (example) may not show correctly for a given time period because that same feature has a different graduated symbol (blue, 50%) for a different temporal period. So I think the slider is working but correctly but the correct features are not displaying in the web map because there is no respect of draw order....my working theory anyway.
To test my theory, I filtered the records per attribute (time period) to see if my suspicion is right....and yes, it seems to matter. if I filter by a certain time period (AM for example), my time slider looks correct as in ArcMap and shows only those features with the but only for that temporal period. And if I also look at popup order.....its all over the place so presumably ArcGIS Online does not use any sort of draw/popup order.
So is display order to blame? How can you ensure that the correct features for your temporal data display at the correct time? Its too farfetched to think that showing traffic patterns by time of day (example- more cars in AM, more trucks in PM, in the same count locations) should be a problem or super hard to visualize. What am I missing? Or, is this another inconsistency between Desktop and Online? I am struggling to find a solution and this is as far as I have come, even with support cases. What choices do I have besides butchering my data merely to visualize? thanks-