Jump to content

User talk:Happy5214/New WikiWork

Page contents not supported in other languages.
From Wikipedia, the free encyclopedia

It's an interesting idea that raises some questions. –Fredddie 22:34, 2 May 2014 (UTC)[reply]

  1. Can I see your math? (Pick an easy state for brevity's sake)
  2. Is the WW divisor still 6 or is it scaled to 6?
  3. Is it easy to game?
  4. Do we need to conduct an importance audit? (Ugh.)
  5. Importance assessments have been subjective in the past, do we need to make them objective?
Here you go:
  1. For Guam:
    1 Mid-importance GA = 1*2*1 = 2 points
    1 High-importance Start = 1*5*1.5 = 7.5 points
    17 Mid-importance Starts = 17*5*1 = 85 points
    94.5/19 articles = 4.974
  2. I'm not sure how I could scale this. The hypothetical maximum is 12, for a task force composed entirely of Top-importance Stubs. But I've noticed that the majority of articles are Mid-importance, and there's really no multiplier in effect for those. To be honest, I didn't think the results would be this comparable at all.
  3. No system is foolproof. The only change from the existing WikiWork formula is the importance multiplier. Therefore, it has all of the inherent limitations of the original formula.
  4. Yes, but we need to do one anyway. Perhaps we can do that during our next assessment audit.
  5. Normally, I would say, "Absolutely!" I'm usually all for objectivity in assessment. But this system rests on the premise that higher-importance articles are a higher priority when it comes to article improvement, and adjusts the WikiWork accordingly. Therefore, articles that are high priorities, like our most viewed articles, should be assessed with a higher importance.
Hope that helps. My plan is to make this a part of a suite of alternative, almost sabermetric, WikiWorks, each tailored to specific goals within our project. -happy5214 07:10, 3 May 2014 (UTC)[reply]
I wonder if the multiplier should be flipped. I was going to talk about the ratio we were talking about earlier today when I realized it doesn't take FAs into account at all. On the WW scale, FAs are worth 0 since they can't get any higher. So based on your Guam example, if that Mid-GA was a Mid-FA, the score would be 4.868. So I'm thinking the multiplier should be reversed, that way a golf score (low is good) is still the goal.
So re-scoring your Guam example
1 Mid-importance GA = 1*2*1.5 = 3 points
1 High-importance Start = 1*5*1 = 5 points
17 Mid-importance Starts = 17*5*1.5 = 127.5 points
135.5/19 articles = 7.131
I think 7.131 says a lot more than 4.974 does. What do you think? –Fredddie 19:32, 3 May 2014 (UTC)[reply]
I agree that the importance numbers should be flipped to keep the golf scoring intact. Also, I think some objectivity is needed in the importance ratings. Some editors have used NHS status to bump an article up a level, yet the last highway bill really expanded the NHS by a lot. I think if we used the new NHS to increase importance, we'd over-inflate things. How I see it:
  • Top = system articles, like Interstate Highway System, Michigan State Trunkline Highway System, etc. An IAR to these guidelines would be things like US 66 based on page views and our desire to get the Mother Road improved, but such exceptions would very be limited.
  • High = any other articles meriting an increase on page view stats in a limited fashion or list articles for things that would be Mid (List of Interstate Highways in Michigan, etc)
  • Mid = most state-maintained highways, regardless of NHS status; lists of low-importance items
  • Low = state-maintained special routes, individual and most county-maintained roads; any lists of bottom-importance items, if applicable
  • Bottom = I don't know that we're using this one, but if we did, I'd split Mid and Low to be primary and secondary state systems and use Bottom for county roads.
Thoughts? Imzadi 1979  19:54, 3 May 2014 (UTC)[reply]
I'm not sure where I'd put articles on individual interchanges or individual junctions, but I'd say they went into low- or bottom-importance. Imzadi 1979  19:55, 3 May 2014 (UTC)[reply]