As most of you know, climbing has made it to the Big Show that is the Olympics. Most also know that climbing is being given one medal in a combined event consisting of speed, boulder, and lead events, with an athletes rank in each being multiplied to produce a final score, low score wins. We’re not going to get very deep into the format here, but instead focus on the peculiarities of the multiplied-rank scoring. If you want to argue about the event format, there are plenty of other outlets on the internet for that.
This multiplication method has similarities to the old USA Bouldering “geometric average” method that was around a season or two….. before they went to 5-10-15-25 zones…. before they ditched that for the IFSC “zone” scoring…. which they used previously (kinda). Reading our breakdown of that might be worth a minute of your time, but we’re going to focus on the Olympics in this one.
The first thing to note, is, you’re at a huge benefit if you have a unique skill (*cough* speedclimbers *cough*). Doubly so if your competition has overlapping skills and splits the higher positions in the remaining rounds, like, say the overlap in boulder and lead athletes. It’s imperative to win (and I mean win – 1st place) where you can, and I’ll prove it mathematically.
In the finals, across 3 events, going 8,8,1 will give a score of 64, which whips 7,7,2 who scores 98. It bests 6,6,2 (total = 72). It even ties (but loses the tiebreak to) a “well rounded” climber who’s just above middle each round, with scores of 4,4,4 (total = 64). For any of you that think the “speed specialists” are gonna get slaughtered, I have a feeling if they don’t false start, they’re gonna do much better than you think.
This huge advantage at the top tails off further down the ranks. Lets look at the people in 1st (Adam) and 2nd (Bernd) after the first round, one rank apart. In order to just tie up their scores, Bernd has to place at least 4th, and thats only if Adam is dead last! It’s a little wonky to read, but the chart below shows the possible scores after the first two rounds, and who is ahead. For the competitor who was 2nd in Event 1, there are just 12 (out of 56) Event 2 scenarios where they can pull ahead, and 4 more where they can tie it up.
Let’s look at the other end of the board. Daila was 8th, and Chloe was 7th. With the same table, what do we see? So long as Daila places ahead of Chloe in the next round, even just one spot, she’ll be tied or ahead.
Multiplicative scoring treats 1st and 2nd as 100% different, and 7th and 8th as trivially different (ok, 14%), even though both pairs are just one position apart.
That’s just math, and math don’t bother me (wooo Mathematics Minor!). What actually bothers me most about the scoring method is the impact on the viewer, and the interactions between athletes scores.
Going into the final event, it’s hard to understand what needs to happen for each athlete to win\place. Granted, if the person with the best (lowest) score comes out last in the lead round, means it’s theirs to lose – if they climb further than anyone else, they get a 1, and still have the low score, and definitely take home the gold. Anything other than a top can get a little, or a lot, messy, depending on the other results and margins. Commentators struggle to describe what happens to the podium since there are potentially wild shifts in everyone’s score. Indeed, the commentators at the Pan Am made a few mistakes in coverage, and who can blame them. See below…
What’s more upsetting for me is that as subsequent athletes come out for lead, it can result in re-ordering of the athletes that have already finished. Adam might be ahead of Bernd after both finished lead, and depending on where Chloe finishes, their relative positions might flip flop. Doesn’t make much sense to me that how Mike does can re-decide whether or not I’m better than Tylor, but this is the nature of the 3-event-multiplied-rank format. Just check out this example:
With the first two events finishing as above, the lead round came down to it’s last man out, Colin Duffy. The Americans had a exceptional showing in lead (with Sean topping, followed closely behind by Xander, and Zach just behind that), so everyone was fairly sure one of them was going to secure the Olympic ticket – but which one? In this moment, there are 4 possible outcomes amongst the Americans….
Colin tops and beats Sean’s time – Colin wins:
Colin falls after Xander’s highpoint, but is behind Sean- Zach wins:
Colin falls before Xander’s highpoint – Xander wins:
Colin falls below Zach – Xander wins (tiebreak being Xander besting Zach 2/3 rounds):
That’s right, the final ranks jump around, maybe even going to a tie-break for the top spot, and this is all just considering the placement of one last climber, in the last event, compared to just three others. While some may be mathematically eliminated from the podium, truth is its going to be anyone’s game going into the lead route. While that’s good for viewer tension, its also a recipe for viewer confusion. Especially if the athlete you’re pulling for has edged out their rival, but their berth is still contingent on how later climbers place. Now imagine the qualifying round, where you have ranks from 1 to 20 to juggle around…
******
I’ve gone on record saying this before, but I’ll say it again – I really hate scoring that relies on “rank”. For any singular event, climbing is like other athletics events – everyone can do their personal best, and the rank just falls out of that, without interaction between athletes. Someone throwing a hammer 10 feet further than me might knock me down a spot, but it doesn’t make my throw shorter. Using multiplied rank decouples the actual achievements of the athletes from the results – how well you do (score) is now only meaningful in the context of how the others did.
At the same time, it’s a necessary evil of the combined format. Triathlon is easy, swimming (timed), cycling (timed), and running (timed) – you’re combining 3 things with the same units. Decathlon has an obscure mathematical formula for the “points”, but it takes hard objective data (seconds, minutes, feet, inches) and processes it against a “benchmark” into a result that can be compared over years. Climbing isn’t either one of these; all 3 events are scored differently – seconds on a speed wall, tops and zones for 3 boulders, highpoint on one route. The boulders and routes change every event, so its impossible to create a set of “benchmark” that we could use to back out a decathlon-esque formula (editor: hell we can’t even agree on what benchmarks are for our grades…. I digress). Unfortunately, I don’t have a better proposal for the current combined scoring format – while multiplication gives a little too much weight to the standouts, addition probably doesn’t give enough weight to them. Hopefully, if climbing earns its keep, we’ll see the events split out more officially in later years. Until then, I suggest you all get real familiar with Excel if you want to keep track of the scores at home.
Justin Meserve is a 17-season “competitive” climber who will one day regale these pages with a breakdown of the (6? 7?) different scoring formats USA Climbing has used since he started to pay attention. He also thinks the best climber is the one having the most fun, so clearly the Olympic Gold should go to… well he’s not sure who, but he knows its not gonna be him until he sends his proj.