From Win Ratios to Elo: How Plaid-Patterns.com Just Got Smarter (and Why That Means Less Scrolling for You)

 

From Win Ratios to Elo: How Plaid-Patterns.com Just Got Smarter (and Why That Means Less Scrolling for You)

A follow-up for the designers, fabric artists, and scrapbookers who live in the gallery — and an open invitation for the rest of the internet to join in.

A little while ago I wrote about how the Battle page on Plaid-Patterns.com quietly sorts the gallery for everyone. The system was simple: two patterns, one click, the winner gained a tally, and the rankings page was sorted by win ratio.

Simple is good. But simple had a problem — and that problem has now been fixed. Battle votes are scored with an Elo rating. That's a small change with a surprisingly big payoff for anyone using the gallery for real work.

The problem with "win ratio"

Under the old system, a pattern that won its first two battles had a perfect 2.0 ratio. So did a pattern that won its first one. So did a pattern that happened to be matched up against weaker plaids three times in a row. The leaderboard was crowded with patterns whose only real qualification was not many battles yet.

That's fine when you're just messing around. It's frustrating when you're a quilter on a deadline trying to find a pre-vetted color story.

The deeper issue: a raw win/loss tally doesn't know who you beat. Beating a clearly mediocre swatch should count for less than beating a pattern that the community already loves. The old system treated both wins the same.

What Elo actually does

Elo is the rating system invented for chess and now used everywhere from tennis ladders to competitive video games to dating app matchmaking. The idea is dead simple once you see it:

  • Every pattern starts at the same baseline rating (usually 1500).
  • When two patterns battle, the system predicts the winner based on their current ratings.
  • If the expected winner wins, both ratings barely move.
  • If the underdog wins, the underdog gains a lot of points and the favorite loses a lot.

Translated to plaids: the quality of who you beat matters now. A scrappy new pattern that knocks off a current top-20 plaid skyrockets up the rankings in a single battle. A heavyweight pattern that beats a low-rated one barely moves. The leaderboard reflects strength of opposition, not just count of wins.

Why this saves users even more time

If you came to plaid-patterns.com to find a usable swatch — not to study ranking theory — here's what Elo actually does for you in practice:

  1. Fewer false positives on page 1. No more "perfect 2.0 ratio after two random battles" patterns crowding the top. The front of the rankings now reflects patterns that have survived matches against other strong patterns. That's a much shorter shortlist.

  2. New patterns get sorted faster. Under win-ratio, a freshly uploaded plaid needed a lot of battles before its true rank settled. Elo converges quickly — a handful of well-matched fights and a pattern lands roughly where it belongs. Less waiting for the gallery to find the gems hidden in the latest uploads.

  3. The top is genuinely the top. When you bookmark "page 1 of the rankings" as your starting point for inspiration, you can trust it more. A 1700-rated pattern actually outranks a 1600-rated one in a meaningful, head-to-head sense — not just "got luckier with matchups."

  4. Discovery in the middle gets better too. Elo creates a real, continuous spectrum instead of a pile-up at "ratio = 2." That means the patterns ranked 200–500 are now a great browsing zone for designers who want something interesting but not yet over-used. Scrapbookers especially benefit from this — you often don't want the most popular plaid in the database; you want a deep cut that feels custom.

  5. Less drift over time. As tastes shift (autumn palettes in October, pastels in spring), Elo naturally re-sorts. A pattern that was a top performer last season but now consistently loses to fresh palettes will quietly fall — without an admin ever touching it.

For working designers, the bottom line is the same as before, just stronger: the first page of the rankings is closer than ever to a pre-curated mood board. You spend less time scrolling and more time designing.

Why this is a great moment for the internet to pitch in

Here's the part for the broader crowd — the netizens, the lurkers, the folks who don't make plaids but love a good little corner of the web. Elo turns the Battle page into something genuinely fun to contribute to, for a few reasons:

  • Every click means more now. Under the old ratio system, your vote was a fractional bump on a counter. With Elo, every battle redistributes real rating points between two patterns. You can feel the leaderboard breathing. That's the same psychological hook that makes chess sites and Hot-or-Not-style ranking pages addictive.

  • Upsets are exciting. Watch a small, underdog plaid take down a top-50 favorite, and you've just handed it a huge rating jump. You did that. The kind of micro-stakes drama that makes you want to do "just one more battle."

  • You're helping a tiny, weird, useful corner of the internet stay useful. Plaid-Patterns.com is not VC-funded. It's not running ads at you. It's a public-good gallery that quietly serves quilters, designers, and crafters around the world. Sorting it costs nothing but a few minutes of clicking, and the benefit lands on every future visitor.

  • No account needed, no commitment, no algorithm staring at you. Just open the Battle page and pick the prettier one. Twenty battles takes about three minutes. Those three minutes shift real rating points and improve the gallery for everybody.

A nudge for the curious

If you've ever sent a friend a link to a "this or that" voting site and watched them lose half an hour to it, you already know the format works. Elo is what makes it feel like a real competition instead of a popularity contest — and that's exactly the kind of small mechanical upgrade that turns casual visitors into repeat contributors.

So: bring a friend. Share the Battle page. Tell them they're literally rewriting the rankings one click at a time. Crowdsourcing taste only works if the crowd shows up, and Elo finally gives every member of that crowd a vote that counts for exactly the right amount.

Three minutes from you. A better gallery for everyone tomorrow. That's a deal worth showing up for.





Comments

Popular posts from this blog

Center Objects Based (COB) Morphing of two images

Banana Peel Luck with Lottery Small Wins

Music Styles Tags List/Randomizer