Skip to main content

Best 10-turn rolling average for calculating max Ski IQ

Feature RequestsSki:IQ

Use a rolling window to calculate max Ski IQ for the best consecutive 10 turns during a longer run. This will even up the playing field for those that like to do longer runs vs those that try to game the system by doing the minimum number of turns to score higher.

Posted by Robert Hoshino ·
Status: Released22 comments

Log in to comment and vote

Comments22

  • None None

    •

    Jun 26, 2025

    Hey guys, Damian from Carv here, thanks for all the thoughts on this one - This is definitely something that we feel would be super valuable and is something we began looking at in the summer, and will definitely be revisiting!

  • Stuart

    •

    Jun 26, 2025

    Do they only throw out "garbage" turns at the beginning or end? How about in the middle of a run? Like many of us, there are days I ski at crowded resorts and I find there are turns made to avoid traffic and find my way through the pack that I wish would get thrown out. It seems like the turns thrown out should be any that are clearly outliers based on ski IQ for a given set of conditions. Someone who's managed a 140 isn't like to be making intentional turns in the low 100's... so why not toss out anything more than 20% lower than your actual ski IQ? (Or wherever you'd like that threshold to be set...)

  • Daniel

    •

    Jun 26, 2025

    Without this feature, the app creates a perverse incentive to ski in 8- to 10-turn segments and stop skiing when you string together a set of especially great turns. Otherwise, you risk introducing a bad turn or two to the set and decrease your score. This undermines the mission of the app, which is to create a platform for learning improvement.

  • Will

    •

    Jun 26, 2025

    How about having three levels: best consecutive 10, 25 & 50 turns. The 10 turn metric could be the one which is used for overall ski IQ, but users could also view their 25 & 50 turn scores & filter leader boards etc. This would be similar to, for example, indoor cycling metrics for power over different durations (see Zwift!)

    • James Pickell

      •

      Nov 14, 2025

      I like this idea. A lot of apps for bikes track your maximal “power” output… for 1, 2, 5, 10 and 30 minutes, eg different times. The weekend warriors can compete at 1 minutes or 2 but it is next to impossible ke to game the system at higher levels. It also would be a good reflection of your conditioning as to score high at longer runs almost certainly requires you to be in shape.

  • Urszula

    •

    Jun 26, 2025

    With this feature I wouldn't have to stop in the middle of the slope just to get the run splitted in two (e.g. when I want to exclude an uneven part of the slope from the IQ calculation).

  • Zlatko

    •

    Jun 26, 2025

    I wholeheartedly agree with all the comments here, nothing to add. But I would like to put some more weight on this feature request, so to be implemented from CARV team with top urgency! This is the worst aspect of the whole Ski:IQ gamification - spoiling your skiing with frequent stops just to "cheat" the algo. And the fix seems so easy to implemented, that I wonder why still not done?!

  • Daniel Goldenberg

    •

    Jun 26, 2025

    Wholeheartedly agree. It's not a coincidence that the top of the leaderboard is "runs" consisting of 8-10 turns, and folks tailor their skiing for that purpose. Simply scoring the best 10 turn streak on a run would seem to do the same thing without forcing someone to stop, or essentially penalizing a skier for continuing to make turns through the flats at the bottom of a run.

  • Nathan Jones

    •

    Jun 26, 2025

    @David - I agree a longer segment should count for more, but the challenge with longer runs e.g. a top to bottom on the mountain is you're going to get often totally different terrain on the way down, meaning some bits you may get e.g. 100 and others 130 to 150. But if someone is going for increasing SkiIQ, then they can just stop after their great turns, so artificially increasing SkiIQ is built in there anyway. This request just allows people to ski on and enjoy their descent without worrying about stopping for the tech. I'm totally ok for people to do a focused effort of 10 turns to improve their SkiIQ, it's their metric. It's not really for me, but I'm not comparing with others.

  • Mark Minardi

    •

    Jun 26, 2025

    This suggestion comes from one of my posts on the SkiTalk forum. I find the way CARV currently scores the Ski:IQ changes your behavior when skiing the mountain. You have to consider where to "define" a segment within a ski run and then you have to just ski at least 8 great turns in that segment. This doesn't allow for building into your best turns. The "rolling window" suggestion would take your 10 best CONSECUTIVE turns from any segment. In this way, all CARV users would be scored on the same number of turns (no more gaming the system by only skiing 8 perfect turns) and those turns could happen at any time during your run, or your ski day, or whenever. You no longer need to be concerned about trying to get those turns only at a particular place and time and having to abruptly end the segment to avoid getting "bad" turns into the mix.

  • Rudy Defruyt

    •

    Jun 26, 2025

    Using rolling median would be better then using average. With median some bad turns are not overly represented like in average calculation