Oh and 65444 is a ridiculously tough and interesting Round 1. Didn't think there existed a Round 1 I hadn't already "solved".

I've probably used ~80 seconds per game on average, basically clicking buttons and seeing what happens. So our approaches are very different! But I think I can play mostly pretty well even when playing fast (and related to timed a couple of days ago: I don't think there is a big dropoff even when I'm drunk). I was pretty happy after first 5 games, stopped there. Then hours later made a mistake of playing the games 6-10 with like my C game, the strategy tip I posted itt pretty well reflects my mental sharpness at the time of playing those games (happened in another tourney played during same session of games). Being tired has a big effect on how well I'm able to think/play the game. So if you need your A+ game to beat THOSE games, I must be pretty far ahead of you.

This is wrong, the situations are FAR from identical. Do you realise I would've lost regardless of what I did? Your play is correct given you gave up on 6s early. My play was correct for how I had played, not giving up on 6s early, and having one 6 in center zone as a result. We both made the correct play, you won the following 60-40 flip.

So was this a recent update, possibly related to BR? Or have I somehow not noticed it in the past.



Has anyone done anything for the BRI spreadsheet? Kokiri? I think I have an idea on how to possibly make it work automatically (would only have to c-p scores), but don't want to waste time on doing that if someone has already done it, or knows exactly how to make everything work without having to try. If not, I can probably try to see if my idea works later today.

I've not done anything, and I probably won't for a day or two, but I might well regardless of whether someone else does. I doubt it would take me much effort to implement something fairly effective.

Yesterday it looked like I could still only see my own scores, which is part of the reason that I've not looked at it yet.

The other question is what alternative scoring systems people want to implement. I've not been paying very close attention to this thread.

I'm only taking around 2-3 minutes per game. It's just that if there is a tough spot, I will pay much more attention to it than I would in another game.

Games 11-20 are definitely not my A+ game.


I was mostly just happy I made the correct play. HOWEVER; your play is correct given my board as the other play gives you roughly 34% equity (68% draw with no 61 in Round 9). What if I had played differently? What if I had the same board as you? What if I hadn't made the correct decision in Round 8? Then cancelling 1s becomes correct by a big margin.

I'm not sure if there is a way to have automated scoring for ties directly. Probably need to use some indirect method for that.

I guess the plan was to have both the lenC DC scoring and one point per player beaten scoring. With ties splitting points between tied players.

In before I lose those 2-8.

I'll try to play them tomorrow. Too tired atm to give you a proper challenge. Just completely butchered the one game I played, or maybe more accurately tried playing.

I can't know what you have done. If I had actually paid attention I could've guessed the most likely scenario, but I hadn't. I just saw I had the 6 in center, and thought I should go for no 1s last rack. It was correct, at least this time.



I'm a bit annoyed at the BRI games lagging. It must be due to the sheer size of the tourney having lots of data, so there probably isn't anything to do about it. Just not ideal. Have to learn to deal with it.



And lastly, lenC.

Crabshoot D under way!



Seems like the total scores for BR test run are NOT working correctly. I wonder if the same will happen in BRI too.

Confirmed. I didn't play especially well either, and was pretty shocked to win some of those.

I also think I made a mistake in game 12. The way I had failed with almost everything in prior racks, I should've 4-zoned 5s in rack 8. I was actually gonna do it at first, but then undid that plan to make the wrong play instead. Oops.

I'm not entirely sure my play was actually that bad, but I'm pretty convinced I had some horrific run-bad. There are a couple of comical rounds in there.

Wow. I won "Battle Royal Test Run" decisively. I had no expectation of doing so after playing the games.

Yet, I was only 4th in total score.

I'm not100% sure what the desired treatment is for ties, but you can duo something with rank bottom up and rank top down, I think. My working draft is something like [rank_up - 1 + (23-rank_down-rank_up)/2]

How does lenC dc scoring work?

Assume this is due to taking my time in the test one, sorry everyone. Can't really play much these days so I stopped signing up for every regular tourney. I think I'll get back to my normal routine next week, I'll pick up the pace vs Kokiri and in BR then.

Whichever system you use, yes you have to average the points and hand out the fractional ones. I ignored this because if 5 people split a DC win, the software still hands out full wins instead of 0.2, this is not the case here.

Kokiri, "my system" just hands out points to top10 players(25-18-15-12-10-8-6-4-2-1) regardless of how many people played. I feel like this is top heavy enough for us commoners to have a shot(maybe not over a 100 games but idk) yet takes away the absurd variance of just counting wins.

What Paul proposed makes lots of sense on multiple levels but I doubt it would be as exciting to follow. Not sure how to hand out the win with this one but we'll have a much better grasp on these things once we play the first one out.

Something like that will probably work for the 'point per player beaten' system, yes. Since point value per position stays constant. But that does not work for the lenC scoring where the points you get from tie scores differ depending on position and how many players are tied. I'm really not sure if there is a good way to make it work tbh. I kind of failed to consider this earlier when I said these would be similar from techical pov.

It was, but I wasn't too serious about it. You were given 30 days to play, so you were free to take all the time you wanted. I mostly wanted to see the full results to use them as test data for the alternative scoring methods.

We should also have scores for first 50 and last 50 games for BRI, maybe even split it to 4 parts? Dunno, is that too much? But this would be to have some intermediate targets too. And to have people who are drawing dead to win overall after 50 games (like me ) to have something to play for in the last 50 games.



Hmm, the road to nowhere lobby seems to have a problem, first 16 games are hidden for some reason.

.
Pretty confident I can excel crush LenC scoring, too given a little time to think about it. For ties, you share the points for the relevant ranks? Might be a little tricky for three way+ ties, I guess.
Just assume I lost most of them.

Yeah, figured it out:

donk has sneaked to 4th place in DC rating

So say there is a 3-way tie for 1st, this would give all of those 19.33 points? Not sure how, but if you say so.

Have you coded the spreadsheet? If so, I can do testing with the test run scores if you either a) upload it to google drive and pm me the link, or b) can send the file to me by email. If b, I'll pm you my email.

Your pony, it is slow.



lol me, time to stop playing for today.

There's a big table that says an n-way tie for mth place scores p points, and the big grid just looks up there based on the player's rank for that day and the number of players with that rank. Not implemented it yet but it should be fine.

I'll play with it a bit and send it to you somehow. I'll not have given it much serious scrutiny for errors but hopefully anything will come out in the wash.

Sounds good!

I'll send you a pm.

If there is something that needs to be fixed, I can probably do that myself. If not, I'll send you a list or something.

Getting to a new rating high and then losing 8 points in 4 days was fun. Hopefully today makes up quite a bit.

Yesterday more than 30 players did the DC

Not sure if that is a record, but at least it is not very usual.

It's far from a record - 42 player one from 2011, think I've seen high 40s as well.

Unfortunetely that yesterdays one only featured 2-3 guys that have started recently so it was just a case of every regular playing.

There are 7.246.052.948 people on this earth that have won less Triplechain Daily Challenges than me and 0 that have won more than me.

Children have posters of me hanging in their bedroom, men want my signature, women are throwing their underwear at me.

My life is now complete

Idea: Battle Royal setup, each round the weakest score gets eliminated(maybe rand when there's a tie so a single one goes). It's pretty beta stuff but people would really need to alter whatever their goto strategy with regard to DC/tourneys/timed is so might be interesting.

Hard to pull off but maybe with capped signups and/or 24h time per round, it'd be doable.

Congratulations! This is the first time you've had the solo lead right?

I had this idea a while ago but it seemed no-one liked it. Although it was before tournaments existed so maybe was just too much work to implement.