If you are like me, you scratched your head throughout the weekend wondering what the heck these “points” were that were being used to differentiate between competitors with a similar number of tops.  The camera lingered on Dan Woods confused face for what seemed like ages as Chris Danielson tried to explain it.  That said, I’m not going to judge Dan – I have a Masters Degree in engineering, and it still took me a few reads of the rulebook to understand what was going on, and more importantly, what it actually means.

Actual wording of the rules is here, Section 8.4

First, the What:

1st differentiator, as it should be, is tops.  You complete more than the rest, you win (See: The Puccio Method)

2nd differentiator is points, and this is where it gets tricky:

Lets assume there are 6 competitors, but this can be used for any number.

On any given problem, competitors are assigned a “rank”.  Best case scenario, each of the 6 fall off at a different spot.  Highest gets 1 point, next highest gets 2 points, so on and so forth.  In the event of ties, competitors are given the “average” rank.  So if there are 2 at the highpoint, each gets (1+2 / 2) = 1.5 points.   A three way tie for 3rd (on a given boulder) would result in all three receiving 4 points.

Confused yet?  In short, at the end, for each problem, each competitor has somewhere between 1 and 6 points.  That means for 4 problems, you have 4 ‘ranks’.  Just add them up, low score wins, like golf, right?  Wrong.

Your final score is the nth root of all your ranks multiplied together:
Total Points = nth√ (RankB1 * RankB2 * RankB3 * RankB4)

This is the geometric mean (average) as opposed to the arithmetic mean, which more people may be used to. Great post on that by math wiz and boulder crusher Matt Wilder found here.

Now, the USA climbing rulebook is fuzzy in 2 places here: what ‘n’ is (I think we can assume its the total number of problems, but its not spelled out), and how exactly the “rank” is determined (high point, attempts to highpoint, do flashes matter, etc etc).  UPDATE: Based on the score sheet posted for last nights finals, rank is by high point followed by attempts to high point.

In the end, everyone has a number of points between 1 and 6.  Whew.

Now the Why:

Why make it so complicated?  The new system rewards competitors for doing consistently better than the others, across multiple problems.  Perfect example of when I think this system would have been better was during a IFSC World Cup event last year.  Akiyo Noguchi flashed 3 of 4 problems, and struggled with a crux start move on the last problem, racking up a handful of falls.  Anna Stöhr topped the first 3 problems in 2-3 goes each, and then, by topping the last problem second go, had 1 less “attempt to top”, with an equal number of tops, beating out Akiyo. This new ‘points’ system would have given Akiyo more credit for besting Anna on 3 of 4 boulders.  It also takes the sting out of problems where either everyone tops, or everyone gets stuck at the same stopper move. 

The new system bumps the reward up for someone who can actually do a move nobody else can.  This is how Mohammed won this weekend – being the only person to top a problem earns you a rank of 1 for that problem. If everyone else gets stuck at the crux highpoint, they all get a  rank of 4. This is a huge advantage in the overall. Josh-Smash Larson getting higher on Problem 4 faster (2 attempts vs. 3) actually hurt Dan Woods’ score; had Josh not gotten so high so fast, Dan’s points would have been 2.3 to Jafari’s 2.7, as opposed to 2.8.  (UPDATE: The real kickers here were P3 and P4 – Jafari was the only on to top P3, earning a rank of 1, Dan tied for worst on P3, giving him a rank of 6.5.  P4, Daniel was ranked 2nd, and Jafari 4th. The points are completely relative to your fellow competitors, and not directly to the problem itself.) 

Downfall here is for morpho problems. Lets say there are 2 short guys, and one tall guy.  Tall guy makes the dyno on P1, short guys don’t.  Tall guys rank is 1, short guys ranks are 2.5. Next problem, sit start.  Short guys send, tall guy can’t pull on. Short guys get 1.5, tall guy gets 3. Each have 1 top, but the short guys have 1.93 points and tall guy has 1.73 points.  Tall guy wins. You are at an advantage if you can exploit a strength nobody else has, even more so if your weakness is a strength of multiple others.

I’m not completely sold on the new system. The most obvious reason, is that its hard as hell to follow from home without a calculator and USA Climbing coming out with an explanation of exactly what goes into deciding the “rank”.  I think intuitively it also makes it feel more like a direct competition (sport vs. game, coming to a Jams-my-Cams soon), where it feels like you benefit more directly from someone else doing poorly. You can get a better “rank” by virtue of someone else having a bad night/attempt.  While, globally, this has always been true, I prefer the notion that each competitors climbs happen in a completely independent bubble, and we just tally at the end, may he who had the best night win.

-Justin Meserve.  A committed competitor and likes to exploit every advantage he can.  Basically, he can crush without having to try.