cookie monster:
I surmise that the 50-move draw rule in chess was based on empirical data from games played at the highest level and player input from masters and grandmasters.

No single game or match would have been a reasonable test of the rule. I personally think the best test of such a draw rule for Espionage would be one in which many of the highest level players evaluate a variety of game positions and determine the least number of moves without a capture to advance towards a win for each position. Those positions from which no clear advantage can be consistently developed should be excluded. Even if the advantage is not always in favor of the same side for a given position the result will be included if the position consistently leads to a win.
From the data of these positions, the number of moves without capture to bring the game to a win should tabulated. The maximum number of moves (from the various positions) will not necessarily be the threshold for the draw rule, but likely will be a smaller number. That is, as in chess, though some positions may be winnable, the excessive number of moves to accomplish it with solid play on both sides will justify calling it a draw by rule. Ultimately, the number of moves for the draw rule should be based on what is usually reasonable for top level players.

I decline participating in the test suggested by cookie monster.