All of these are extremely important!


1) Allow us to make 0 picks in manual distribution! Sometimes we just want true randomness.

2) Give the AI a spot on the ladders! Let it play in the seasonal & 1v1 ladders at the very least, so 300 people can mock it regularly ^_^

3) When people play on the tutorials, why are there no links to the written guides the community has written? This is a pretty massive oversight, and tbh it doesn't look like the FFA/1v1 auto game community does a good job of educating players on actual skill. And the single-player tutorials teach you just enough that you can finish turns.

4) Among other problems, the realtime ladder has a huge bias towards a good start. For average players who eventually become good, it's simply not possible to make a meaningful enough difference in their mean rating, which will be forever stuck due to past performance. The only solution is firing up alts, which I assume isn't what Fizzer had in mind...? TrueSkill favors the first generation and punishes people who start off worse than average. If your first 50-100 games are bad, you'll always look bad to the eyes of the realtime ladder.

5) Make clan league written into the website itself. You owe that to the clan community. And do it right! Don't mess around the way you've done with the realtime ladder templates, that falls into the "among other problems" in #4, and I'll be writing a followup on that later.

#4 is nonsense.

Buns lost his first 8 games in a row. He lost over half of his first 50 games.

4 doesn't seem to be the case for TrueSkill; the algorithm will widen sigma for a while if it detects you improving but once your new skill level becomes consistent, sigma should narrow again.

Backing 1, 2, and 3.

He was already an above average player by that point, though. Could you really claim that a bottom-ranked player today, with 100 games played, could ever have any hope of reaching the top without making an alt?

Maybe I'm just not seeing it.

Edited 11/22/2016 03:07:49

Agree totally with #4. The rating system on the RT ladder needs to be scrapped and replaced, or fixed

The best players were never "bad". They were inexperienced, they were still better than average for the amount they'd played even then.

The RT ladder encourages runs less than 1v1 and 2v2 ladders do.

And no buns wasn't already "good" when he joined the RT ladder. He was rank 57, with 1255 rating.

He wasn't a bad player then, but he was an inexperienced player. He joined Warlight October 1st 2014, he joined the RT ladder March 10th 2015, 5 months after he began playing. He didn't do well till much later.

Some of those games I had no idea what I was doing. I really only played ror and ffa's before trying that ladder.

It's slower if you have a bad start, but it isn't impossible by any means.

1 - try automatic distribution.
3 - see https://www.warlight.net/wiki/User:DanWL60/Single_Player_Levels_2016. Doesn't quite cover all levels. Looks better in dark-theme btw - https://www.warlight.net/Forum/220284-darktheme-wl-wiki-small-styling-fixes.

Edited 11/22/2016 10:49:27

how about the #1 suggestion on uservoice? It has 1000 votes, and given that everyone has 3 votes and some people have alts voting, there must be at least 300 people that voted on it! There are a lot of players who only play diplo/non-competitive games, plus the competitive ones, who would play that! It would revitalize diplo games.

https://warlight.uservoice.com/forums/77051-warlight-features/suggestions/1088481-peace-treaties

Uservoice is where ideas go to die

Speaking of alts, I too sign Fateless' petition!

Next petition:

On the ladder pages, why can't we search for specific names? It's a pain to find someone who has hidden their ladder stats on their profile.

I'm starting to wonder how long Fizzer's roadmap is... and how much of it is tarmac and how much is gravel.