Let's evaluate the position. First, if Barry rolls a 5, I am all but dead. I still have two men to escape and bring round whilst he is nearly home free. If I manage to release one man the other is likely to be attacked and possibly closed out. The worst scenario is that both men get closed out and I lose a gammon. As the position is highly volatile, this certainly looks to be a good double. Remember Woolsey's law: if you're not sure whether it's a double, then double anyway!
But ... 25 times in 36 Barry will stay on the bar. Then if I can roll any 6 or 42 or 52 (even 32 and 21 are quite good) and Barry stays on the bar again, I could be looking at redoubling to 16. How often will this happen? Looked at in percentages for each roll, 70 per cent of the time Barry will stay on the bar, 42 per cent of the time I will roll a 6, 42 or 52, and 70 per cent of the time Barry will stay on the bar again. The likelihood of these three events happening in sequence is the product of the three percentages: about 21 per cent.
Other sequences will certainly give me my required 25 per cent to take. Should I fear the gammon? Not really, the chance of both my men being closed and then my losing a gammon is fairly remote (7 per cent according to Jellyfish). So, all in all a correct double and an equally correct take. Sadly, over the board I got it wrong and passed. Fear of the remote gammon got to me. An excellent pressure double by the fearless McAdam, which reaped a rich dividend.