When I first flew down to Rockingham to help Sam and Bill, I asked around for someone who knew the current DA. When no one had one, I brought my own unit down to the next meet at Valdosta. A big problem with using the prediction function is the accuracy of the data being used. Any abnormal inputs will alter the estimate. So, if you get out of it on the top end, you can't use that input. If the DA drops and you rejet, the old data will be off some for future runs. That's why I still like the log books, which gives you quick access to previous runs under similar conditions. It has worked well for me over the last 15 years. My current meter is about 5 years old. It's been back for recalibration once to the factory. They drift some on DA over time. I think that Bill will confirm that I still remember a little about ET predictions. We dialed Sam for a 2 to 4 hundredth cushion. I was a little further off at Rockingham, but didn't have any data records to use. The only time that Sam got dialed really tight was when we were called up to the lanes and went right out onto the track without the time to go back to the trailer to find dial in numbers. The 12.4 dial I thought would be a first round breakout; with a buyback likely, I didn't worry too much. Sam got a 12.403 and a round win. I didn't have enough confidence with the limited data to dial it much closer than .02. If you have the data and the confidence in the bike's consistency, you can run .012 over in the first round and up your dial-in for the second round by .01 and run .001 over for that pass like I did in a Prostar meet. When you're dealing with thousandths of variations, you will go rounds. Yes, I'm talking about my Hondamatics. The clutch CB750 bikes just can't repeat that well. I'll post some time slips next season to support my beliefs. Many of the Hondas running at Valdosta cut good reaction times, but didn't last too many rounds. Either the driver didn't dial it right or mechanical issues affected the run that put them out of the event. Building it just before the race, going to the race without practice, is only good for a timeslip not a lucky win light. When you can consistently run sub .05 packages, you will become a serious threat to win some $$