There are two schools of thought with probe placement. The describe method is one way it keeps the interior clear and is a easier set up, it also leaves an air pocket around the UTH (glass likes even heat not uneven localized heat) that could be an issue. All t-stats set up like this will vary when the room temp changes the probe is not exposed to the room temperature there is no way for it to compensate for it. If you have stable unchanging room temps this is an excellent choice.

The other method is to place the probe inside the tank directly over the UTH on the glass bottom. (DO NOT USE TAPE!) This method allows for ambient air temps and will keep more accurate temperatures in rooms that change temperatures a lot. (I had a 14ºF drop over the last 14 hours for example)

The PROBE MUST NOT MOVE!!!!!

In both set ups the probe must not move it is critical. The external placement there is possibly less actions that can move it but if it moves it is an emergency. The internal placement the snake can pull it and still it must be held in place. I personally use hot melt glue to fix it down I can pull it off if needed and re glue it and return the snake when it cools. easy.

That being said I always recommend a back up t-stat to add a measure of safety to any set up, especially a interior placement. The 500r has one good thing about it it is cheap, you get what you pay for. I don't believe it to be trustworthy it is difficult to get set correctly and then even when it is it has a long throw between the temp it is set to and the temp it shuts down and starts up with. It can make a ok failsafe but agin setting it is a pain. My advise is to find something else. Sorry to say this but the difficulty in setting the unit make it such a poor choice that I find it unusable.