Request for alternative scoring systems

[hozn]

I like freeze points and welcome more charting ideas, especially if they include volunteering to write the necessary queries! (Which is what vvill did for freeze points, I should note.)

I think FS has “worked” with its current scoring, whatever that means exactly, but I don’t think it would fail if the points system made the game more dynamic or more rewarding of riding in gritty conditions rather than turning out 200-mile days.

I know there are a number of veterans that aren’t doing it this year because they would rather not deal with the commitment. It does make me wonder if there might be a better system. But the one we have has been fun for many people.

[Steve O]

I don’t know how to write queries, but here’s how I might devise a “crowd-sourced” bad day scoring bonus. The tricky part is that it’s dynamic across the contest; that is, your older points can change as the game continues.

Assumption:
The fewer people who ride, the worse the conditions. Given there’s no perfect way to determine how bad it is, I think this is probably the best proxy for determining the combination of weather and road/trail conditions.

– Rank the days based on number of riders
– the 10 days with lowest rider numbers get more highly weighted points like this:

• Worst day – 2.0
• 2nd worst – 1.9
• 3rd worst – 1.8

etc. to 10th worst – 1.1

What makes this tricky is that come March, if there’s a horrific rainstorm that makes March 11 the 3rd lowest riding day, then it retroactively changes the points for the previous days 4-10.

(variation – 20 worst days: 2.0, 1.95, 1.9, 1.85, 1.8, etc…….)

Maybe someone taking a SQL class can take this on as a project.

[jrenaut]

I don’t think I’m in favor of retroactively changing the scoring. I like your idea, though.

[hozn]

I like Steve’s idea. Even enough to do the SQL work if there is interest in using it. Personally, I don’t think retroactive scoring is terrible, since we already allow people to sit on rides (not upload) for the entire length of the competition, if they so choose. Maybe we could tighten the window a bit, though, so that the worst days were only per week or something. So that you would have a pretty good idea before end of competition.

I also think the freeze points method (if I remember right) of tapering off awards for mileage so that the value of miles drop off above a certain number might help keep the pressure on team members to all do their parts. And maybe we could have the pointless prizes earn real bonus points for the teams?

I do think it would be interesting to keep the leaderboards more dynamic and diminish rewards for churning out 300-mile weeks in favor of more reward for participating in the games, group rides, etc. that probably make this more fun. This isn’t the NBC. Plus it is sad that Subby isn’t playing this year.

I do think we should provide extra points (double points?) for team group rides … (We can track those in Strava.)

[vvill]

I also like Steve O’s suggestion – why not just do it off a baseline, say 200 riders/day = 1.0x multiplier

then just skew it as much as we like, say 100 riders/day = 1.5x or 50 riders/day = 2.0x or something. Whatever really. Either linear or some more hyperbolic/trigonometric function. Whatever chosen baseline of 200 riders or whatever doesn’t really matter since every day is scored against the same baseline, and the query will be a lot easier. Scores may change obviously as some rides are logged later on, but ultimately credit is given where it’s due.

This could maybe be combined with diminishing returns for longer rides. Then we don’t need to worry about actual weather data and sampling at the beginning or middle or end of the day.

[ewilliams0305]

@vvill 150880 wrote:

This could maybe be combined with diminishing returns for longer rides

This sounds like a conspiracy to make sure I suck! While I understand what your trying to do and I understand the spirit isn’t just to go out and ride big miles. I also KNOW what it’s like to literally ride outside in the cold for 20 hours below freezing. Anyone who’s ever done this surely wouldn’t want their points to be diminished. It’s possibly the hardest thing anyone could ever do on a bicycle period.

I’d love some background into why we are even discussing theses changes. We’re people actually unhappy with the current scoring system? In fact the current scoring system has allowed people who don’t ride a ton to be the most valuable team members. In fact the daily 11points has always been what makes or breaks a team. In fact the teams that have always won have been the teams with the most rider daily participation, not miles. In fact isn’t that the point (to get people to participate), so haven’t we already created the perfect scoring system?

Sent from my iPhone using Tapatalk

[Judd]

@ewilliams0305 150883 wrote:

This sounds like a conspiracy to make sure I suck! While I understand what your trying to do and I understand the spirit isn’t just to go out and ride big miles. I also KNOW what it’s like to literally ride outside in the cold for 20 hours below freezing. Anyone who’s ever done this surely wouldn’t want their points to be diminished. It’s possibly the hardest thing anyone could ever do on a bicycle period.

I’d love some background into why we are even discussing theses changes. We’re people actually unhappy with the current scoring system? In fact the current scoring system has allowed people who don’t ride a ton to be the most valuable team members. In fact the daily 11points has always been what makes or breaks a team. In fact the teams that have always won have been the teams with the most rider daily participation, not miles. In fact isn’t that the point (to get people to participate), so haven’t we already created the perfect scoring system?

Sent from my iPhone using Tapatalk

Fact: Eric must be destroyed.

My understanding that the official scoring will remain unchanged. While I found Freeze Points to be silly last year and will continue to find any system with complex mathematical formulas to be silly, I’m also generally for adding as many additional metrics to the competition. Since most people aren’t going to be willing to make the extreme sacrifices necessary to be the top rider, the additional layers of data allow for competition within the competition. Last year my competition was to have the team with the most Hains Point laps and also to finish above a team mate with a similar riding pattern. In my mind, I won Freezing Saddles last year without diminishing the brutality that the folks at the top of the official scoring system put themselves through.

Sent from my iPhone using Tapatalk

[rcannon100]

@ewilliams0305 150883 wrote:

I’d love some background into why we are even discussing theses changes.

Boredom.

Jealousy.

Envy.

Because some of us are confused and think that all this matters.

We all want to be like you – we just dont want to actually do the work.

Laziness.

Because every year Shawn Dread is on my team.

Riding in a bike in the cold sucks and we are generally trying to change the game to avoid this.

Consensus rules (a.k.a. socialism)

[jrenaut]

My intention is to see if we can make more teams competitive longer. I don’t want to penalize anyone for riding, but how many teams had even a distant shot at winning after the end of January?

Also, if you look at the leaderboards, there’s almost no change if you rank by points or miles. So if the ten point bonus barely changes the standings, all it does is motivate people to ride. What if we could do the scoring to motivate AND affect the standings? That is what I’m looking for.

Sent from my XT1060 using Tapatalk

[LhasaCM]

@jrenaut 150887 wrote:

My intention is to see if we can make more teams competitive longer. I don’t want to penalize anyone for riding, but how many teams had even a distant shot at winning after the end of January?

Also, if you look at the leaderboards, there’s almost no change if you rank by points or miles. So if the ten point bonus barely changes the standings, all it does is motivate people to ride. What if we could do the scoring to motivate AND affect the standings? That is what I’m looking for.

Sent from my XT1060 using Tapatalk

Then for this alternative approach, is it a multiplier that should be applied to mileage, or just applied to the 10 point “participation” amount? For example, if you added a multiplier to the 10 point bonus based on some function of that day’s participation percentage (just based on the ratio is probably too simplistic), you’d be rewarding people who ride on days when other people don’t ride by an equal amount, regardless of how far someone goes. That would increase the incentive to participate on any given day.

Or, so that there is some reward for additional mileage, as previously mentioned you could make any multiplier/bonus logarithmic so there are diminishing returns (e.g., the more miles, the more points…but the marginal gain for the 5th mile is more than the 100th).

The devil, of course, is in the details. The challenge with any scoring system is trying to identify what you want the outcome to have been, then jury rig something that would’ve resulted in that outcome in the past, and crossing your fingers that it works OK the next time around.

(Disclaimer – as a newbie here and a “on a good day I’ll have a 5 mile round trip commute” type of rider…I’ve got no strong feelings about what is “right.” Just intrigued by the mental gymnastics of scorekeeping )

[vvill]

@ewilliams0305 150883 wrote:

This sounds like a conspiracy to make sure I suck! While I understand what your trying to do and I understand the spirit isn’t just to go out and ride big miles. I also KNOW what it’s like to literally ride outside in the cold for 20 hours below freezing. Anyone who’s ever done this surely wouldn’t want their points to be diminished. It’s possibly the hardest thing anyone could ever do on a bicycle period.

I’d love some background into why we are even discussing theses changes. We’re people actually unhappy with the current scoring system? In fact the current scoring system has allowed people who don’t ride a ton to be the most valuable team members. In fact the daily 11points has always been what makes or breaks a team. In fact the teams that have always won have been the teams with the most rider daily participation, not miles. In fact isn’t that the point (to get people to participate), so haven’t we already created the perfect scoring system?

jrenaut replied already, but yeah basically the intention was to have an alternative scoring alongside the current scoring system that would keep riders’ and teams’ scores closer, and also to reward just regular riding (especially in bad conditions) more than 8 hr+ rides. This different scoring is meant to encourage regular riding in any conditions, more than just monster mileage and monster hours, which some riders (including veterans) don’t want to commit to. I don’t think the current scoring system was ever going to be replaced.

In any case, I’m not particularly invested in whatever is used (I signed up for slackers). I do find it fun/interesting to play around with the data.

[jrenaut]

Interesting idea to scale the bonus more than the mileage.

The cool thing is we can put up a bunch of leaderboards side by side and see if there’s one that works much better than the others, and then maybe we discuss using that one as the official scoring next year.

Sent from my XT1060 using Tapatalk

[rcannon100]

@vvill 150889 wrote:

I signed up for slackers.

Go! Team! Woo!

[rcannon100]

Can you track indoor trainer miles? I know they don’t count. But that’s the point. How perfect a pointless prize then for a metric that does not count?

[cvcalhoun]

@jrenaut 150887 wrote:

My intention is to see if we can make more teams competitive longer. I don’t want to penalize anyone for riding, but how many teams had even a distant shot at winning after the end of January?

Also, if you look at the leaderboards, there’s almost no change if you rank by points or miles. So if the ten point bonus barely changes the standings, all it does is motivate people to ride. What if we could do the scoring to motivate AND affect the standings? That is what I’m looking for.

Sent from my XT1060 using Tapatalk

If the idea is just to make teams competitive longer, we should have smaller teams. The first year I played, we had teams of 9 or 10, and it wasn’t until the last week that my team pulled far enough ahead so that the second place team couldn’t catch up to us. My understanding is that the year before that, the teams were even smaller, and the standings closer. Once the teams get to a certain size, the point differences between them get large enough so that no one individual can say, “Oh, I’ll go out this weekend and ride a century to make up the gap.” So you don’t see major changes over time in which team is ahead.

[Author]

Posts