In HMS based heat racing, you need to plan and expect around 3 heats an hour, lets say 20 minutes a heat.

Everything you do for a heat needs to fit into that 20 minute window. Wen we started trying to Parallel Score, it was surprising to find that everything could fit into what anyway seemed a busy 20 minute window,

When Parallel Scoring on your own, the first secret is to choose a Parallel Scoring system where the work basically happens at a different part of the 20 minute cycle. If I’m on my own, I enter scores into AFleet at the line looking over the Line Judges shoulders. I love it actually. Let’s say that happens in the first 2-3 minutes of the 20 minute cycle. You can’t basically get going with HMS Excel until the Baord Manager or Line Judge hands you the score sheet around minute 4. Entering a large heat carefully into HMS Excel takes only maybe three to four minutes. The rest of your time is spent with backups, publishing results and “issue management” plus queries.

What happens at a larger championship where you have more than one person to help. Check out our heat scoring workflow diagram above from the Worlds. Once you have more than one person you can immediately start to tackle the key objective of ERROR FREE RESULTS.

This concept of AUDITS in teams-driven parallel scoring is HUGE.

At the Worlds our scoring process was built around two big audits for each race cycle. A Worlds is high volume, believe me. One audit (top of the diagram) happens every heat, Audit 2 happens at the end of each race. Heat sheets will be handed over by the Line Judge/Board manager, and if you’re lucky, Umpire decision paperwork arrives at the same time.

Our process at the top of the chart, Data Entry, was that a QA Scorer would interpret and call the results from the sheet – and one scorer on each of HMS and Afleet enter the results as they are called. Then the scorer on the HMS Primary system would call the sail numbers back to the QA Scorer and the AFleet scorer to check everyone was in agreement and correct. It’s interesting the numbers of things you pick up – even when you think you’ve been careful.

The big focus is to keep errors from ever entering the system in the first place. Trap them at data entry. I’d say that 1-2 errors per day were caught this way at the worlds. It’s a very human process – so you get handwriting legibility, keying errors (nerves, workload, rain etc) and you certainly get umpire judgements that need triple checking and discussing. Stop errors getting in.

Do your backups heat by heat. Honestly it’s a big piece of work. It takes longer than entering results. JFDI and don’t miss a heat. We used the Andrew Crocker Fleetboard method for our backups and will describe that in another article.

If you are dealing with low order heats simply go back and do the next one as it arrives. 20 minutes flies by as there are always “issues” that need handling – if only hydration or lunch!

If you have just entered results for an A-heat, there is more audit work to do. Our process was to get the QA and Data Entry scorers assembled and check output two ways. The number one objective is to not let results containing an error out of the room. The HMS Primary scorer would print the results ready to go, then do two sets of calls to the other team members:-

1/ Call down the total computed scores column

2/ Call down the position column by sail number

If you can think of something better, let me know. But calling that set of 168 numbers takes around 3-4 minutes. They are some of the best 3-4 minutes you can use.

Honestly when doing it, you get desperate to reach the end of the two calls and everyone says “I agree”. If that doesn’t happen, it’s an emergency clinic session to find the problem and fix it.

Only when the HMS Primary Scorer and QA Caller says the results are good to go, does publication to the Official Notice Board, Facebook and Website happen.

Every time we officially released results we knew that both scoring systems were completely aligned and error free up to the completion of that heat. How often can we normally claim that? Those results were 100% correct. Every time.

Go read the flowchart above and contact us via this website if you have questions….

Autopsy:-

Well, you should be thinking…. “If you’re using one scoring system and not parallel scoring, what happens normally??”

You know, I’ve been worried about that too.

I suspect as a matter of course, it’s not uncommon that Regatta results come out with little winkles buried in them. I’ve heard it said, “the answer to your question is whatever the system says…”

So I think we ‘usually’ see regattas with imbedded errors. We must do.

If the Scorer detects an error though, instead of cursing we should celebrate. In this high volume data, human driven environment there will be errors. Just find them. It’s good.

At the IOM Nationals, our error checking was only half as good as we did at the Worlds. We had around 1000 Nationals boat results in each of the systems and actually our audits successfully picked up two errors… in a thousand results. One on each of HMS Excel and Afleet as it turned out. I dont think anyone would have noticed, but the point is that we found them and sorted it. Two errors per thousand boat results. Celebrate finding them.

But at the Worlds, no results sheets left the scorers office until they were clean. Honestly. Those results were perfect. Every time.

I was interested as to what the formal Worlds process was for an aggrieved competitor to query their score. Honestly it was a very protesty environment …. I thought scoring queries were inevitable. I hadn’t seen it before but the Umpires had a formal “Scoring Enquiry Sheet” that a competitor with a scoring grievance had to fill in to send to the Scorers.

During 90 heats of the IOM Worlds, we received precisely zero Scoring Enquiry Sheets.

Although we felt emotionally that the final 2-way audit before publishing was the big leap forward, I’ve had time to reflect. In retrospect, the challenge is to prevent errors ever getting in to the systems in the first place. Audit 1 in the diagram above is probably the key. Together as a Team – Check Umpire decision completion very carefully, review handwriting squiggles, boats in the wrong heat (if any) and get that data extremely clean before entering it in your systems.

It can still be wrong – I feel badly now that the Nationals Race Committee handed the Scorers a verbal Redress Decision for boat 95 (!!). I know now that the Race Committee was wrong and as Scorers we should not have let the error in. It is actually OK to tell the Race Committee they were wrong too – it simply takes courage!! I must apologise to Graham when I next see him 🙂

So – fully audited results cross checked across two systems… is this a world first in radio sailing?

Probably.