The time of Event 18 is not the trial end, either. If you do not know this yet, please read this manual page.
https://monkeylogic.nimh.nih.gov/docs_GettingStarted.html#AlignTimestampsAndAnalogData
In NIMH ML, each trial starts at Time 0. Events 9 & 18 are stamped sometime during a trial just like any other code. If the timestamp of the 9 is 1.234 ms, for example, it means that the 9 was stamped 1.234 ms after the trial starts, not that the trial started at 1.234 ms. Data acquisition also starts at Time 0, so it is wrong to say that the data was recorded between Events 9 and 18.
Those two built-in codes were introduced by the original ML. I did not removed them in NIMH ML, because people already had scripts looking for them and I wanted to keep the compatibility. They will be the first thing to be removed though, if NIMH ML 3.0 comes out.