should olympic clocks go the extra decimal/s to eliminate ties?

[Goose]

The two man bobsled (if you don't already know) tied gold for Canada & Germany. Skiing has not been ever without a tie as well and while I don't get to follow it, Id assume these things may have happened in world cup competitions too. And I can recall occasional medal ties in other sports through the years not just for gold but also sliver and bronze. Rare things but does happen.

But anyway it got me thinking should they go to the next decimal or not? I kind of like the tie for gold (or any medal) to be honest. I think it holds good sportsmanship when teams of the same medal congratulate each other and two different countries share the same glory moment together. But I can also see how there should only be one winner of each. I mean that's what sports at the highest levels is. Its a winner (or not) and a second and so on.

But if they start going from 100ths of a sec to thousands of a sec can we really say one beat the next if its only by 1-1thousands of a sec? And what. if that ties? Do we then go to the 10 thousands of a sec? We actually have the tech to do that. Though I'd have to think with the rare times this happens now that if it went to thousands it would never happen and if it did then so be it.

But there is also something to be said for conditions between first runners vs last runners too and certainly when precipitaition may be falling in an event so there is all kinds of things that may or may not be totally fair anyway. But a clock is a clock and at least for that there is nothing to skew it. So should they bring the extra decimal or not. As saidI like the idea of sharing something all that close. But if it were me and I beat you at something Ive worked all my life at by a thousands of a sec then I might feel I should get that win I deserve. So I honestly don't know what the right answer should be.

 

[François Pugh]

I think the 0.01 second clock is good. A tie once in a while is ok.

 

[crgildart]

I saw one of the speedskating events they had a tie at hundredths and were able to drill deeper in to the clock to thousandths to break the tie. I think with the cumulative aspect of bobsled they would have had to go back and take all of everyones' other runs also to the thousandths then re add them. It's not as simple as just looking at each of their final runs at that level. with cumulative runs added.

I guess it depends on how often after three runs added up there is still a tie down to hundredths? Sounds like a solution looking for a problem most of the time.

 

[pchewn]

An occasional tie is fine. If it starts to be a recurring theme, then it is time to go to the next decimal place. When you get to the .001'' measurements, you are talking about the difference in the start wand being hit with the ankles near the tip of the wand vs 1 inch closer to the pivot point. Is that really what we want to distinguish between gold and silver?

 

[mdf]

When I was young, the Olympics bragged about "swiss timing" -- analog. As technology improved, the resolution of the clocks got better and better. Eventually, the various sports decided finer resolution was meaningless and froze the number of decimal places.

Also, the clocks may be accurate to a millionth of a second (or not - are they kept in a temperature-controlled enclosure?), but that does not mean the timing system is that accurate. How reproducible, at the microsecond level, is the signal from the mechanical switch in a start wand? What is the diameter of the optical beam that you break at the finish line? Does blocking 80% of it generate a trigger in some atmospheric conditions, while other conditions require blocking 99% of it?

 

[Eleeski]

In the days of slide rules we had to carefully keep track of significant figures. Once digital measurements became common, the understanding of decimal points got lost. A clock is capable of displaying to .001 but it may not be capable to discern a true .001 difference. I don't consider @pchewn 's wand hit location a real difference. That less than an inch difference associated with .001 must be properly calibrated. And @mdf has good insight about the accuracy of the whole system.

Note that the start must be accurate as well - it's not the same as a photo finish. The gun goes off at the same time for everybody in a bunch start race - reaction time is a skill worthy of determining a winner. The less than an inch gap visible in a photo finish might not be real in remotely timed setup.

Accept the ties that are within the tolerances that really must be there for honest measurements.

Eric

 

[BGreen]

In most cases, alpine racing is timed to 1/10,000th and truncated to 1/100th. Electronic timing can be accurate to 1/1000th. @mdf brings up the point of total system accuracy and repeatability. Even if the the gear is accurate and repeatable, it doesn’t mean the wiring is. Further if the wire gets cut and spliced mid race, then what? I thought it was an interesting decision when syncs changed from syncing the systems from the wand to simply synchronizing the clocks.

So yes, most clocks are precise to 1/10,000th. I don’t know how accurate they are because there is drift. 1/100th is a small but meaningful amount. 1/1,000th IMO isn’t.

However, changing timing to 1/1,000 would necessitate a rule change. Backup timing is done by hand — two people with stopwatches. That is frequently accurate to 1/100th, beyond that is not possible from my experience.

 

[pchewn]

The gun goes off at the same time for everybody in a bunch start race - reaction time is a skill worthy of determining a winner.

Eric

This brings up an interesting point for me. The skier-cross is a real race (all people on the course at once, group start, 1st to the finish wins). Whereas what we call "Ski Racing" (SL, GS, DH, etc) is not a race at all. It is a speed event where the times get compared from one person to another.

In the case of skier-cross, we don't even need a timing device, just a way to see who finishes 1st, 2nd, 3rd.

 

[tball]

We should probably be more concerned with timing systems being hacked.

 

[Goose]

interesting thoughts in this thread. I don't think the accuracy is of issue as for being able to time to the 1/1000 of a sec and in fact I believe it can be precise to even more than that. (or should we say "less" than that..lol). Atomic clocks are precise yet even they are affected by temperature but we are talking millionths of a second so certainly we could go thousandths and even 10thoussandths of a sec without accuracy issues if they really wanted to imo.
But even at just going thousandths, is it something desired? IDK I suppose if ties were more common they might consider it. But I also think we can reach a point were records keep breaking and it has to come to a place eventually where that must reach a limit. I mean we cant spend no time racing, we cant finish soon as we start. So there has to eventually be a limit as finishing faster and faster is not physically possible anymore or also for safety reasons. Therefore possibly leaving us with only thousandths of a second to determine things anyway. Are we there yet? probably not but may not be too far away.

 

[Primoz]

No. It's fine the way it is. In xc skiing, they actually eliminated hundredths of second in individual start races after Wassberg won Mieto for 0.01sec in 15km race (at that time about 40min race) at Lake Placid Olympics. After that time is measured to 0.1sec and if times are same it's tie. For mass start and for sprint races, where it's head to head competition, there's always photo finish and 1cm can make difference between winner and second placed one. But for individual start, I'm actually fine with 0.1sec.

 

[oldschoolskier]

In multiple lane events (ie swimming) I would say no. Timing lag in different cable lengths would start to show up and may or may not cause errors. If you are interested look up Velocity Factor in cable and the effect it has, yes it measurable in length of cable. Add into that kinks changes in VF and connectors if not installed correctly. Add in temperature change this changes this number as well (think super cooling to increase clock speeds on computers) as accuracy goes up so the effects of outside conditions. Will every race be run in identical -10C condition top to bottom +/- 0.001 degrees.

For single lane events (applies to multi-lane as well) the photo finish that is being used currently, will likely not confirm results as distances would not be measurable given camera locations.

In on event at they did run to 0.001 of the timing on display however not sure if this is what was actually used.

Remember the accuracy of timing or distance is only as good as its weakest link and there are lots from actually timing to conditions changing. Either. We allow variance and sacrifice and compensate with a little accuracy or if we increase accuracy let’s ensure that everyone competes in the “exact” same conditions. Fair is fair.

So for the foreseable future I would still recommend 0.01 accuracy.

 

[fatbob]

Stick with 0.01. It's already an amount of time we can't really contextualise so no need to separate performance by more.

 

[jonc]

I do timing for our local race team so have some experience in this. Our rules are handed down from the FIS competition rules. Electronic timers are capable of recording times down to 1/10,000 and there is a synchronized backup electronic system and synchronized hand timers. Accuracy is incredibly important, we have to power on the systems early enough before the race to warm up and stabilize, sync happens as close to race start as possible, resync is performed before each run. We check for time drift before and after the race. We only use times to the 1/100th of a second and ties are allowed, they do happen sometimes, we actually had 2 in our last race though not for podium spots.

All that said, I think 1/100ths are precise enough for a 1+ minute race, it means ties will occur but there are enough variables that it is ok for the competitors.

I also saw on the speed skating there was a tie in one of the short races (~30sec). They broke the tie for TV viewers by going down to 1/10,000th of a second. Even for such a short event this seemed excessive. In the end the tie did not affect the medals anyway and officially the results showed a tie.

 

[crgildart]

If we need to measure down to that level of precision to ensure fairness, then perhaps we also need to ensure the competitors have the exact same course conditions to compete on too?? I'm kidding of course, but as long as we're picking nits there are much bigger things that aren't truly fair to all competitors.

 

[mdf]

So there are approximately three questions here
- should we use the next few decimal places on the existing timing system? An obvious no, for the accuracy reasons several people have talked about.

- could we build a timing system accurate to a few more decimals? Obviously yes, but it would be expensive and inconvenient to install, calibrate, and maintain.

- and the opinion piece, would it be a good idea if the technical means existed? My opinion, no, let a tie be a tie. The appropriate time resolution varies depending on the length and type of the race, and how controlled the course conditions are. (I was not aware of the 0.1 second resolution on XC.)

 

[KevinF]

This is an interesting discussion and involves some variables I hadn't considered (i.e., the consistency of the system, etc).

At any rate, my local night league race (about a 25 second course) which operates in a dual panel slalom format does time to the 0.001 second. By our league rules, beating your opponent is meaningful in terms of standings, etc. Last week alone we had two races decided by less then 0.01 seconds. I've lost (and won...) one or two over the years by 0.01 seconds or less as well (and losing is never fun, but losing by an eye blink really sucks).

I always figured that if our local beer league is timing to the thousandth that "real" races are timing to closer precision then that. i.e., there's what they're timing to, and there's what gets displayed on the TV.

 

[Goose]

In multiple lane events (ie swimming) I would say no. Timing lag in different cable lengths would start to show up and may or may not cause errors. If you are interested look up Velocity Factor in cable and the effect it has, yes it measurable in length of cable. Add into that kinks changes in VF and connectors if not installed correctly. Add in temperature change this changes this number as well (think super cooling to increase clock speeds on computers) as accuracy goes up so the effects of outside conditions. Will every race be run in identical -10C condition top to bottom +/- 0.001 degrees.

For single lane events (applies to multi-lane as well) the photo finish that is being used currently, will likely not confirm results as distances would not be measurable given camera locations.

In on event at they did run to 0.001 of the timing on display however not sure if this is what was actually used.

Remember the accuracy of timing or distance is only as good as its weakest link and there are lots from actually timing to conditions changing. Either. We allow variance and sacrifice and compensate with a little accuracy or if we increase accuracy let’s ensure that everyone competes in the “exact” same conditions. Fair is fair.

So for the foreseable future I would still recommend 0.01 accuracy.

I agree conditions play a big role. A lot can happen to any snow/ice surface between first run and last and can vary in between as well The sun can go in and out, the temps can change, precipitation may or may not fall, and wear of the surface itself, etc...
That stuff can certainly have a positive or negative affect on each persons time and be far more than any thousandths or even hundredths or perhaps even tenths of sec. But the difference with that stuff though is that it cant be helped. Its just an ugly nature of things and often a matter of luck. But keeping time (if one wanted to make the point when compared to that other stuff) can be helped. You can do something about that if you wanted to. Not suggesting it should be done but just to say it can be.

 

[Philpug]

Ties should come down to rock-paper-scissors-lizard-spock or two man potato sack races on concurrent Sundays.

 

[crgildart]

Ties should come down to rock-paper-scissors-lizard-spock or two man potato sack races on concurrent Sundays.

Came back to post that same comment.. How about a Hippity Hop elimination final?