Discussion: Surebridge Finishing Leg Discrepancies?

I was surprised by the difference in time between my electronically timed, identical finishing legs on Sat and Sun (28 and 19 seconds) as I felt like my run-in speeds were fairly similar. I was even more surprised by the discrepancy in John Fredrickson's times (15 and 4 (!!) seconds). I doubt it made any significant difference in the final results, but these data certainly suggest that there was as difference in synchronization in the various finish punches, or possibly some other issue in the timing units to account for these results. If it is helpful to the analysis, I do specifically recall that I punched at different finish punches on the two days. I wonder if somebody more knowledgeable than I would be able to comment on this issue in the interests of not reproducing this error in more striking (and perhaps consequential) fashion at a future meet.

Were there one or two GO controls (#100)?

At least two.

Nadim noted this discrepancy and talked to the crew about it, I believe. I think he said one box was replaced partway through Saturday.
If you look at the splits, they're all over the place.
Hey Valerie, is there some way they could figure out the time difference?

Late Sunday GO control photo. Apparently one was red-shifted and the other blue-shifted.

there was only one go control on day one and three finish units. day two had two go controls. I requested an additional one as I lost a second waiting for the unit to free up on day one.

The finish line crew did put out a new control. I'm not sure but seem to remember them saying it was a replacement. They recognized a synchronization discrepancy (around 8-10 seconds?) and thought they had corrected this on Saturday. On both days I had punched the right most of the three controls at the finish line with times of around 20 seconds. I did take it easy on the finish leg on Saturday but ran it fast on Sunday. The crew offered to correct it if it were to make a difference in placing.

The SI units (BSF7's and 8's) still drift enough (as much as 10 seconds in a week) that we resynchronize the controls right before every meet using SI Manager. I have also found that synchronizing controls using SI Manager sometimes takes as many as 3 synchronize requests per control to get it right on. So it was taking about 20 minutes to do 25 controls.

We just have started using the BSM8 Timemaster system and this seems to sync them much quicker and right on first time (as well as the added benefit of clearing the control memory of previous races).

I would recommend a) the Timemaster approach and b) syncing right before placement for every event (to be sure to be within 2 second on all units) if > 2 days since previous syncing. Otherwise there will be problems with not only splits but end to end times (if there is more than one start punch (if punching start used) and/or more than one finish punch)

Seems to me that all start and finish punches could be synced immediately before the event (i.e. in the morning), which covers the most important timing issue. Splits are less critical because they don't affect the final results.

The actual reason is hell hole #2, a small edition of the Surebridge mine hell hole right at the finish. HH have the effect of a black hole in space. Eddy can tell you more about those, he is the space expert.

Now, to be serious: Before you guys go all over with speculations, the timing crew knows about the problem and works on correcting it. Allow them a few nights of good sleep. Some of us subsisted on 2 - 4 hours of sleep last week before the event.

I check with them to find out what the facts are.

I'm fine with waiting and am not concerned. I hope others are too. I enjoyed the meet regardless of the place that I finished. Unless they know which boxes were put out when, it could be very hard to straighten out. I only just now compared this: My official and hand timed duration matched on Saturday. The finish split was 21 seconds. My official time was 5 seconds faster than my hand timed duration on Sunday and the finish split was 20 seconds.

The full technical explanation is on the way from the R&D department of HVO's timing kings. - They can query the SI units to see who used which box when. So it is going to be settled and the updated results/splits should reflect the changes.

I'm sure I used different units on the two days (0:30 and 0:13 last splits), but my overall times are within 3 seconds of what my watch says for both days. If the problem is only with the splits to and from the last control, but not with the overall elapsed times, I think it is not worth the EP crew's time to fix it. Splits are there to give an idea, not an end in itself.

(Speaking as someone who endured six hours of pouring rain and 6 °C under a mt. laurel bush at Surebridge on a certain day in October, 1993, taking the damn splits by watch, pen and paper.)

I introduced this inquiry not in the interest of correcting some minor variations in the results of the Surebridge event, but rather as a means to avoid similar problems at future events. My sense is that the others who have contributed to this thread have the same goal in mind. I don't feel the need to change any results, and would certainly discourage a laborious manual adjustment of the finishing times, but I certainly do wish to clarify what steps should be taken to avoid (or at least limit) the types of inconsistent data which appear to have been an issue this past weekend. In other words, I hope the HVO timing kings can devote their energies to educating those of us ignorant souls who plan to introduce electronic timing to our local clubs in the next few months on how to avoid duplicating this type of systemic aberration.

Only an astronomer would come up with that red-shift, blue-shift line. Laughing out loud! Way to go, Eddie.

It's that old joke... "Why did you run the red light?" "Sorry officer, I was going so fast, it got blueshifted to green"

Here is the response from the R&D department of the Timing Ministery:

"Three finish boxes were used on Saturday. The finish crew asked for a fourth finish box on Sunday to handle the White and Yellow runners approaching from the opposite direction of the more advanced courses (in a new finish chute). During their initial setup (before the fourth finish unit was ready), the finish crew placed two of Saturday's finish boxes in the advanced finish chute and one of Saturday's finish boxes in the new White/Yellow finish chute. The new finish box was programmed from one of the results computers, and then added to the advanced finish chute by the finish team. (This new box was the one on the competitors' right as they approached the finish line.)

When the new finish box was programmed, its clock was synced up with the computer's internal clock. We had used an SI Master Unit to sync up the times of all of the control units on the previous Sunday evening, and again on the Friday morning before the event. This SI Master Unit had been synced up with a PC that had its clock set from an internet time server the Sunday before the event. We should have used this same SI Master Unit to set the clock of the new finish box after setting it to be a finish box, but forgot about this step. (Apparently there was a small time difference between the clock in the SI Master Unit and the clock in the computer that was used to program the new finish box.)

After hearing about the timing discrepancy for the finish leg, the finish and e-punching teams visually compared the times on the three finish units in the advanced chute. This was done by looking at the LCD on the bottom of each box. (Aside: These were "older" control boxes. The newest units have the LCDs on the top.) The LCD cycles through displaying several different things. One of these is the time. It took a little bit of interpolation to determine the time difference because the units do not all display the same thing at the same time. (E.g. One may be showing the unit's programmed control number while the other shows the current time, so it may not be possible to view the time on the different units at the same time.)

We discovered that the clock for the box on the right (as the runners approached the finish from control #100) was just under 10 seconds faster than the time on the other two units.

We used the SI Master Unit to set the time on the new finish box so it matched the time on the other two finish boxes.

=== Lessons learned ===

1) When programming a new control unit, don't forget to synchronize the time. Verify time synchronization after the box has been programmed.

2) Whenever more than one control is used at the same location (most commonly at start, finish and GO), the clocks should be resynchronized before the first runner of the day starts. Synchronization should be verified by the Start and Finish crews by a visual inspection of the LCDs on the control units." (from Stephen Stibler's e-mail to me)

We used our e-punching system the first time at an A-meet. But since all the timers are computer workers in their day job, they put quite some effort in developing the system and ironing out the kinks.

They already presented one minor adjustment that I really liked: The split times are listed top to bottom instead the tedious left to right lay-out that is commonly used.