Big Figure Days

Interested to know if anybody accounts for figure releases in their mechanical systems.
For example.
A system like JonnyT's Spot On would probably take a short this morning on the downside break in EURUSD.
Is this a good idea ahead of non-farms or is it prudent to wait until the release before doing anything?

Naturally if the figure is the "right" way around it would be advantageous but then this "if" possibly makes things more of a gamble than a trade?? Maybe over time it evens out each way..I don't know, has anybody done this sort of figure analysis?
Personally I have not factored figures into any mechanical system but I certainly take great account of it with the scalping, ensuring I do not have any potentially dodgy orders before or through any major announcement.

FWIW I have looked at the non-farm payroll days results for the past 13 months (not a huge sample I know) for the JonnyT type break out system & GBP, 70 pip stop. At first glance results don't seem to be anything out of the ordinary ie -38, -22, NT, +56, -70, +54, +70, +103, -70, NT, -70, NT, NT and today NT. Bearing in mind that this is with a 2 hour time restriction on entry, so often the markets may be waiting for the news release and nothing much happens early on. An all day entry study would be more interesting (will post later), as would FOMC meetings etc.

I've been looking at this sort of thing recently. I have a mechanical support/resistance system for cable and I was curious to see if it would improve if I kept out of the market around announcements.
So far I have just programmed in (in Tradestation) the 1:30 pm announcement dates so that if there is an announcement that day I won't enter a position between 1:00 and 2:00.
I haven't so far filtered the announcements as to whether they are (considered) important or not - I just included all of them.
The result is (obviously) a reduction in the total # of trades but a modest increase in the average trade size and PF. My conclusion is that the improvement is signifiant (i.e worth doing) if not dramatic.

This is for a short term supp/res scalp strategy though. I haven't tried it on a breakout system that might have a longer holding period.

Thanks for the feedback.
Rog - those results are interesting as if you stayed out on non-farm days you would have ended up about square (+13) so maybe it does just cancel out over time. think it would be interesting to see on non farm days if there is a reversal or continuation of morning price action. That will provide the answer for any breakout system regardless of timeframe. Something like (Close- 13:29hrs price) + (13:29hrs price - Open) .
Jim - do you just enter each date seperately into EL code in TS or is there a way to reference some sort of calendar?