But he had two groups in the 1994 experiment, the low-RWA group and the high-RWA group. Both played the same game. There's no evidence of different stimuli at play there. Yet different results. Where do you draw the conclusion that the study treated the two groups differently? That's not backed up by anything in the source.
In the 1998 games, they did two games. One with just RWAs, but they scored low on the new SDO measure. And the second one was RWAs, but a single high-scoing RWA/SDO planted at random in each region. Other than that, these two games were run identically, yet they also saw marked differences in how events unfolded.
In the first 1998 game (RWAs only) it ended up being isolationist with very little inter-regional communication. There were no conferences, virtually no trade and no negotiations, and regional leaders basically focused on just their region. There was clearly distrust, but nobody actually stepped up to lead a war. In the second 1998 game, every single SDO ended up self-selecting to become some sort of political power-broker in their region, and they bargained like crazy with the other elites, to get more power. It also ended in an arms-race ike the 1994 RWA game, but the game's turn limit ran out before that could resolve.
For both the 1994 and 1998 games they played two games each with the same rules, but with different composition of players, and very different results. So I don't think you can say the game was engineered to give one specific result.