Interesting that you don't count things like Powered Coasters or Big Apples in your steel coaster poll and that Tyrolean Tubtwist at Joyland in Great Yarmouth isn't in the wood poll this year despite being included in the past.
When the poll was done with text ballots, I was using a tabulation program that I wrote after teaching myself how to code. It was crude at best, but it worked. In the second year, Grant Barker came on board to take over the coding bit. He improved it (on our end - stuff the voter doesn't see) massively and we started talking about a steel poll. Creating and managing a text ballot for a steel poll was much too large a task for just the two of us and I seriously doubt many voters would be willing to slog through a text-based ballot with 2500 entries on it, anyway.
Several voters had asked for an easier way to vote and a few had suggested using spreadsheets so they could re-sort their ballots by ranking as they went. We opted to go that route and Grant got to work converting the tabulator into CSV-based rather than text-based. With that change, Grant came upon the idea to write a program that would use the RCDB database to auto-create a spreadsheet ballot. Just run the program, put in your filter parameters (such as only listing coasters that were open this year), and voila! There's your ballot. This made a steel poll possible.
HOORAY! Except that the steel poll had 4000+ entries. That's nuts.
So we started talking about how to reduce the size of that ballot to make it manageable enough that people would actually fill one out. We decided early on that powered coasters and alpine coasters wouldn't be included, but the ballot was still way too big. That's when decisions got harder.
We considered Mitch Hawker's option of "your fave [type of coaster]" where you simply rank your fave Batman clone, your fave boomerang, etc. I shot that down early on, because some people really struggle with knowing what particular category a coaster fits into ("Is this a Zyklon or a Galaxi?" "Vekoma corkscrew or Arrow?") and sometimes, one or two of those coaster models are substantially better than their clones. Take boomerangs, for example. I hate them. Yes, even the ones with the vests. Once and done, if at all.
Except the one at Wiener Prater. I love that one. I'll pay for re-rides. I would easily rank that in my top 50, while the others would be in the bottom tier. Do I want to have a "fave boomerang" line that would make me decide between artificially inflating the status of 20 or so crappy boomerangs just to get Prater's a solid ranking, or do I punish Prater's version to make the others' rank more accurate?
So rather than do that, we opted to eliminate some of the stuff that wouldn't factor into the rankings much. Let's face it, once you get past the top 50 or so, most people couldn't care less how something ranked. They want to see the best 50 and the worst 10 or so. Big Apples, Wacky Worms, butterflies, kiddie coasters, etc... they'd all be stuck in the middle bit that everyone would just scroll past. Out they went.
So even with those omissions, the ballot is still 2500+ lines long. It's still tedious, but not heinous. So far, most of the voters are filling it out and sending it in with their wood ballots. (sigh of relief) A few sent in the wood early, then sent in the steel a couple days later. (another sigh of relief).
So we're pretty happy with the steel ballot this year - once the results come out and we see how some kinds of coasters fare, we might decide to pare it down further for next year. We'll see.
The one part that affected the wood ballot was the omission of powered coasters and kiddie coasters. Tyrolean Tubtwist is listed on RCDB as a powered coaster and Auto Bergbahn at Prater is listed as a kiddie coaster. Both of those dropped off the ballot as a result. Neither of them affect the top 50 nor the bottom 10 rankings.