Open Side Menu Go to the Top

04-24-2012 , 02:33 PM
Quote:
Originally Posted by Gabethebabe
TH10, is there any puzzle where you suspect we have not made the nuts?
Quote:
Originally Posted by TH10
Good question, but I would say it is highly unlikely that there is, I guess it is possible though.
Looks like there was at least one, wp Alex!

Funny that a few days ago I was thinking that I haven't scored 40k+ in DCs for ages, then two in a row. There has never been a 50k+ DC score, should happen at some point.
Triplechain Quote
Triplechain
$25m Guaranteed WPM on CoinPoker
Join the action now
Daily Rewards • Splash Pots • CoinRaces
Triplechain
04-25-2012 , 09:49 AM
SIGH. When I went to bed, 5 hours before the day ended, I was still 5k points in the lead. That's three 2nds in eight days now. But it was also lols, I was genuinely really disappointed with my score when the game ended, then I find out I'm in the lead by miles.
Triplechain Quote
04-25-2012 , 11:03 AM
volutpat causing frustration.

I created several configurations in free move mode that coincided completely with regard to center, corner and edge numbers and I had all the chains and bonus needed.

Only when reproducing the boards, I run into problems.

This board, for example, can not be made, because I run into problems rack 7.

Spoiler:


I have had several other configs that ran into problems in the first 4 racks, and this is the first one that gets past that.

Back to the drawing board.
Triplechain Quote
04-26-2012 , 01:50 AM
Funny

After trying a long time yesterday solving volutpat I woke up this morning and had a new idea. I tried it this morning and solved the puzzle in 10 minutes.
Triplechain Quote
04-26-2012 , 03:53 AM
Quote:
Originally Posted by Paul101
But it was also lols, I was genuinely really disappointed with my score when the game ended, then I find out I'm in the lead by miles.
Well, you can imagine how I felt after playing that DC then, I was sure I would finish in pretty close to last with the score I got, seemed to me like 30k+ should've been quite easy. But I somehow finished 6th out of 26, I have no idea how that is possible, guess everyone made same kind of mistakes in that one.
Triplechain Quote
05-12-2012 , 09:21 AM
well i somehow managed to get DC win number 12 during the blackout. I think it was at least 3 months in the coming, reminds me of how it took me like 6 months to get my first DC win.
Triplechain Quote
05-13-2012 , 12:15 AM
Quote:
Originally Posted by DonkDonkDonkDonk
well i somehow managed to get DC win number 12 during the blackout. I think it was at least 3 months in the coming, reminds me of how it took me like 6 months to get my first DC win.
Pretty sure I shared that win with you, coincidentally that was my first ever!
Triplechain Quote
05-13-2012 , 09:21 AM
Two days in a row where a 2 in Round 9 would give me a win, but it didn't come. So yet more second-best scores. Pretty frustrating!
Triplechain Quote
05-13-2012 , 09:41 AM
Quote:
Originally Posted by zdye724
Pretty sure I shared that win with you, coincidentally that was my first ever!
congrats! First one is the hardest.
Triplechain Quote
05-14-2012 , 03:46 AM
i instinctively choose 1, although if i thought more, i would be tempted a bit by 2.

The problem that strikes me with 3 is that it's hard for it to ever be the best option - you need to roll 3 4s in the last two racks of the second phase AND not throw 2 5s. Otherwise, it's at best only as good as one or other of the two other options. That's i guess the way to decide - look at the possible distributions of 4s/5s in the last two rolls (and roll 9 for option 2), and see which option is right most often. Should be possible to analyse it exhaustively, no

In general atm, I need a pretty good reason (especially late in phase 2) to leave a hanger like the rogue 5 in option 3.

Also: omfgbbqwtfymmv sniped in yesterday's DC
Triplechain Quote
05-14-2012 , 04:56 AM
Quote:
Originally Posted by kokiri
i instinctively choose 1, although if i thought more, i would be tempted a bit by 2.

The problem that strikes me with 3 is that it's hard for it to ever be the best option - you need to roll 3 4s in the last two racks of the second phase AND not throw 2 5s. Otherwise, it's at best only as good as one or other of the two other options. That's i guess the way to decide - look at the possible distributions of 4s/5s in the last two rolls (and roll 9 for option 2), and see which option is right most often. Should be possible to analyse it exhaustively, no

In general atm, I need a pretty good reason (especially late in phase 2) to leave a hanger like the rogue 5 in option 3.

Also: omfgbbqwtfymmv sniped in yesterday's DC
Sorry, I deleted my post, because one of the boards was flawed. I will repost later.

I am continuing my nosedive in the DC rating.
Triplechain Quote
05-14-2012 , 05:08 AM
Here is the problem again. You have this board:


And get a rack of 14566. How do you play?

Option 1.


Option 2.


Option 3.
Triplechain Quote
05-14-2012 , 06:38 AM
1 and 2 are equally valid strategies, 3 seems undesirable.
Triplechain Quote
05-14-2012 , 08:52 AM
Quote:
Originally Posted by kokiri
Also: omfgbbqwtfymmv sniped in yesterday's DC
Lol sorry. I know how it feels to lose it that late; Popcorn did it to me a few weeks ago. Was very shocked to win yesterday tbh!

Also, lol, I complain about running badly with Round 9s then win a DC where I needed Round 9 to contain a 115 and no 2. That's gotta be fairly unlikely, maybe 3%?

Gaby's problem seems trivial: Option 1 all day every day.
Triplechain Quote
05-14-2012 , 10:03 AM
i'd play 1 every time
Triplechain Quote
05-14-2012 , 10:03 AM
Gabe,

I have some text written up about that game you discussed on my work computer, but probably won't be at that computer until tomorrow. It is kind of interesting I think (at least to me) because the differences between the games kind of showed me something that I should take into consideration more carefully.


Quote:
Originally Posted by Paul101
Gaby's problem seems trivial: Option 1 all day every day.
I go with Option 2 here as it gives you a better chance of snaking the 5s.
Triplechain Quote
05-14-2012 , 10:44 AM
Quote:
Originally Posted by Aicirt
Gabe,

I have some text written up about that game you discussed on my work computer, but probably won't be at that computer until tomorrow. It is kind of interesting I think (at least to me) because the differences between the games kind of showed me something that I should take into consideration more carefully.
I think that game was interesting and looking forward to your thoughts.

http://www.triplechain.net/reconstruct?chain_id=1542861

66662 or 66652 in rack 9 gives me standard 60K ...
Triplechain Quote
05-14-2012 , 12:16 PM
1 vs 2:

if you roll 2 or more 5s in rack 7&8 then it's a wash,
if you roll exactly 1 5 in 7&8 then if you roll a 5 in rack 9 you prefer option 2. By my count that's a 40%*40%*2*40% = 11.2% shot.
if you roll no 5s in 7&8 and 1 in 9 then it's a wash
if you roll 0 or 1 in 7&8 but whiff in round 9 then you prefer option 1 (unless you convert back to option 1 with the 5 you get in 7/8). I make that (40%*40%*2+ 40%*40%)*40% = 19.2%

so (modulo my dodgy maths) you prefer option 1 over option 2 19% of the time, and vice versa 11% of the time.

there's also a slight effect on the 4 chain in option 1's favour, since in option one you don't ever need to play in the middle.
Triplechain Quote
05-14-2012 , 01:17 PM
It's hilarious how swingy the game is, about 2 months ago my Daily rating was 10 points behind Gaby, and currently I'm 6 ahead. We keep switching like crazy. Also TH's Daily winrate fluctuations are nuts; he had the hottest December/January of anyone ever and in the last 2 months has won maybe 2 or 3.

Re-revised targets, now I believe I have a shot against TH again:

Combined Rating:

Short-term: Maintain the second-highest combined rating, and get to within 75 of TH.
Medium-term: Get to within 50 of TH.
Long-term: Beat TH.

Timed:

Short-term: Maintain 2nd and Master status in the Timed ranking.
Medium-term: Get to 1610
Long-term: Reach 1625 and pass TH (these will happen around the same time).

Daily Challenge Ranking:

Short-term: Get ahead of Dynasty.
Medium-term: Reach 1480 rating.
Long-term: Beat TH and get Expert status.

Daily Challenge Wins:

Maintain the highest Daily winrate, keep it above 1 in 9 and hopefully get it above 1 in 8.
Long-term: Top the win-list. This is the most-achievable long-term goal, although it's gonna take ages to catch Dynasty; around 3 years at the current rate!

Overall:

Break away from Dynasty and Gaby to become the number 2.
Long-term: Become number 1.

I still don't really care about highscores.
Triplechain Quote
05-14-2012 , 01:33 PM
Quote:
Originally Posted by kokiri
if you roll exactly 1 5 in 7&8 then if you roll a 5 in rack 9 you prefer option 2. By my count that's a 40%*40%*2*40% = 11.2% shot.
This should be 40%*40%*2*60% = 19.3%

Quote:
Originally Posted by kokiri
if you roll 0 or 1 in 7&8 but whiff in round 9 then you prefer option 1 (unless you convert back to option 1 with the 5 you get in 7/8). I make that (40%*40%*2+ 40%*40%)*40% = 19.2%
This should be 19.5% cos of rounding errors.

So roughly the same. But there's also the effect on the 4s, which makes a difference. And you miss out the most important factor, which swings things drastically. I'll leave it to you to figure out what that is.
Triplechain Quote
05-14-2012 , 01:48 PM
given you've got 1's in the 4 outer squares, you really want to be free to put a rack-9 1 in the middle square, which you can't do in the best case scenario for option 2
Triplechain Quote
05-15-2012 , 02:26 AM
Quote:
Originally Posted by zdye724
Pretty sure I shared that win with you, coincidentally that was my first ever!
Congrats!

Quote:
Originally Posted by Paul101
Also, lol, I complain about running badly with Round 9s then win a DC where I needed Round 9 to contain a 115 and no 2. That's gotta be fairly unlikely, maybe 3%?
I did some exact odds calculations once, and iirc the result was that you can get quite close estimations using direct odds multiplication. Not sure if it applies to all situations, but based on it 0.16 for 11, 0.6 for 5, and 0.4 for no 2 = 0.16*0.6*0.4 = 3.84 %, so about 4 %.

Quote:
Originally Posted by Gabethebabe
http://www.triplechain.net/reconstruct?chain_id=1542861

66662 or 66652 in rack 9 gives me standard 60K ...
Bozzer had one normal mode game (http://triplechain.net/reconstruct?chain_id=1516416), where a 66621 last rack would've given 60000 exactly with 25 bonus and One dimensional! That would've been amazing.

Quote:
Originally Posted by Paul101
Also TH's Daily winrate fluctuations are nuts; he had the hottest December/January of anyone ever and in the last 2 months has won maybe 2 or 3.
3, and only 4 in last three months, and also almost 5 weeks without a win now.

Quote:
Originally Posted by Paul101
Daily Challenge Ranking:
Long-term: Beat TH and get Expert status.
The bolded part sounds impossible.

Quote:
Originally Posted by Gabethebabe
Here is the problem again. You have this board:

And get a rack of 14566. How do you play?
Number 3 obv if you know all the last three racks are 55555.

If anyone cares, based on DC stats and the rating calculator it would seem that the average player in DCs has a rating of about 1378, which is lower than I would've expected.

Below are my DC stats for April.

Spoiler:
Stats are listed: January; February; March; April; 2012.

General:
Rating: +20; +1; 0; +3; +24
Days played: 29; 29; 29; 29; 116
Wins: 8 (27.59 %); 2 (6.90 %); 1 (3.45 %); 2 (6.90 %); 13 (11.21 %)
Average players, who finished their games: 30.14; 28.52; 28.52; 26.66; 28.46
Average position (raw positions, ties don't affect this): 7.86; 9.76; 10.07; 8.69; 9.09
Average players tied with: 0.79; 0.72; 0.72; 0.69; 0.73
Head-to-head win % (ties count for this as 0.5 wins): 75.09 %; 66.85 %; 65.73 %; 68.68 %; 69.18 %
NEW: TOP5 finishes without winning (raw positions, ties don't affect this): 7 (24.14 %); 6 (20.69 %); 8 (27.59 %; 13 (44.83 %); 34 (29.31 %)

Finishing positions relative to field (ties count for 0.5 wins):
Beating 0.00 % - 16.66 % of opp.: 0; 0; 2; 1; 3
Beating 16.67 % - 33.33 % of opp.: 2; 1; 2; 3; 8
Beating 33.34 % - 50.00 % of opp.: 3; 7; 4; 3; 17
Beating 50.01 % - 66.66 % of opp.: 6; 7; 6; 4; 23
Beating 66.67 % - 83.33 % of opp.: 4; 6; 5; 6; 21
Beating 83.34 % - 100.00 % of opp.: 14; 8; 10; 12; 44

9th rack luck:
More zones to chains: 6/7 (85.71 %); 7/10 (70.00 %); 4/8 (50.00 %); 7/9 (77.78 %); 24/34 (70.59 %)
Number of games where it was possible to win with 9th rack: 16 (55.17 %); 8 (27.59 %); 9 (31.03 %); 9 (31.03 %); 42 (36.21 %)
Average odds to win with 9th rack for all the days played: 24.41 %; 5.86 %; 8,62 %; 11.72 %; 12.66 %
Expected wins with 9th rack: 7.08; 1.70; 2.50; 3,40; 14.68
Wins compared to win EV: +0.92, +0.30; -1.50; -1.40; -1.68

Very disappointed with the month, since I threw away one win and by my estimation at least 3 rating points because of very stupid mistakes. Should know better than to make those kinds of mistakes. What makes this even worse is that my rating peaked at 1498, so I should've made Daily Master without the mistakes. Rating went 1492->1498->1495, with the last three points lost in the last four days of the month, so another poor end to a month like in March. The month had some other similarities with March too, four bottom third finishes again, which is too much to my liking, also ran more than one win under win EV for rack 9 again, expected wins at 3.4 is actually pretty good. Gaining 3 rating points isn't bad either.

Added a new stat too, to see how many near misses for wins there have been with regards to positions, counting top5 positions without winning, so places 2 to 5, 13 of those in April is very, very high. More zones with 9th rack was above EV once again, but I actually realised that this stat would need context to be actually infomative. We don't know how many chain dice there were in last racks on average, and quite often the amount of chain dice in racks 1-8 are such, that a lot of players will need the same number in last rack to get more zones. So it is very difficult to tell how lucky one has been compared to the field, which is what actually matters with this. Whatever, I guess it still tells the absolute luck for a single player. What is alarming is the amount of players in total, big drop from January to April, can't think of a specific reason for this, but hopefully the trend will turn upwards in the future.
Triplechain Quote
05-15-2012 , 04:04 AM
2x3s and 3x3s in rack 1 tempted me to play 3s in the last two daily challenges.

ugh
Triplechain Quote
05-15-2012 , 04:16 AM
3s are the devil, don't be lured into the trap of their siren song...

Last edited by kokiri; 05-15-2012 at 04:17 AM. Reason: i mean, 4s are pretty borderline as it is
Triplechain Quote
05-15-2012 , 10:23 AM
Quote:
Originally Posted by TH10
I did some exact odds calculations once, and iirc the result was that you can get quite close estimations using direct odds multiplication. Not sure if it applies to all situations, but based on it 0.16 for 11, 0.6 for 5, and 0.4 for no 2 = 0.16*0.6*0.4 = 3.84 %, so about 4 %.
A quick calculation tells me that it's actually 5.7%, so unless I've done something very stupid, your estimate is fairly poor.


Quote:
Originally Posted by TH10
3, and only 4 in last three months, and also almost 5 weeks without a win now.
Grim.

Quote:
Originally Posted by TH10
The bolded part sounds impossible.
Fml, I keep making that mistake :-/


Quote:
Originally Posted by TH10
Very disappointed with the month, since I threw away one win and by my estimation at least 3 rating points because of very stupid mistakes. Should know better than to make those kinds of mistakes.
Don't worry, everyone does it, although very annoying when it happens. I've thrown away a win before by making a very stupid mistake, and last month threw away 2 rating points in one game by misplacing a die. Also remember Gaby's Round 9 mistake which lost the win to Dynasty (cheers Gaby for making my life harder; give it to anyone but Dynasty pls )? If anything, it's proof that you're not a bot!
Triplechain Quote
Triplechain
$25m Guaranteed WPM on CoinPoker
Join the action now
Daily Rewards • Splash Pots • CoinRaces
Triplechain

      
m