1.2 BTC Tournament | Play for free! | 6 Winners | MooCoin RC1

[PendiCoin88]

GOOD LUCK ALL....!!!! #21Hours

[moocoin]

Hi Everyone!

The Release Candidate Tournament is now over!
Congratulations to our winners, and thanks to everyone who was on the site and helped us get the system ready for launch.  We learned a LOT about running the system at scale and that's going to help us make the launch even better for you, the players!

We're going to make some changes to the system before we go live, but they won't take long.  Stay tuned to BitCoinTalk.org for our launch announcement, or check back at www.moocoin.com

Here are the winners of the 6 leaderboards

Most Bitcoin earned through referrals: PUBc4cfa98886fa98146d7cdc3b6342a7a0   - PAID!
Most Bitcoin won: PUB358c2b3af200b823341d4eba16a60122 - PAID!
Best winning ratio: PUBaa46516ff63faaf7daf3aeda9c6a35e5 - PAID!
Most games played: PUB92af0fb5316f5ef9ad340e96d253b4a1 - PAID!
Best winning streak: PUBaa46516ff63faaf7daf3aeda9c6a35e5 - PAID!
Worst losing streak: PUB48cb9dfc3baefb77ba4b1696435c7074 - PAID!

If you won, please email us at admin@moocoin.com as soon as possible to collect your 0.2 BTC reward!

Thanks!
MooCoin

[FUR11]

I'm a little curious as to why my account that went 95-0 didn't win the best win:loss ratio leaderboard.  Seems like 100% should beat 99.68%.

[moocoin]

I'm a little curious as to why my account that went 95-0 didn't win the best win:loss ratio leaderboard.  Seems like 100% should beat 99.68%.

Yep - totally fair question.  We use a weighting algorithm as described in the rules.   More specifically, we use the lower bound of the Wilson interval at a 95% confidence interval to calculate the rankings.  This makes it so that someone who wins 999/1000 games ranks higher than someone who won 1/1 games.  Clearly the guy who won 999/1000 is more likely to be the better player.

Thanks,
MooCoin

[FUR11]

I'm a little curious as to why my account that went 95-0 didn't win the best win:loss ratio leaderboard.  Seems like 100% should beat 99.68%.

Yep - totally fair question.  We use a weighting algorithm as described in the rules.   More specifically, we use the lower bound of the Wilson interval at a 95% confidence interval to calculate the rankings.  This makes it so that someone who wins 999/1000 games ranks higher than someone who won 1/1 games.  Clearly the guy who won 999/1000 is more likely to be the better player.

Thanks,
MooCoin

That makes sense.  I guess my only argument would be that the leader-board doesn't say "best player" it says "best win:loss ratio".  And a 1:0 ratio is better then a 999:1 ratio.  If you wanted to figure out the best player, it seems like it would be some sort of ELO rating system or something.  It's nitpicky on my end, sure, but the whole leaderboard bot stress test system seemed to be built around being nitpicky.

[PendiCoin88]

YEAAAHHHHHHH.........

I'm Winner
Best winning ratio: PUBaa46516ff63faaf7daf3aeda9c6aa35e5
Best winning streak: PUBaa46516ff63faaf7daf3aeda9c6a35e5



Who is PUB92af0fb5316f5ef9ad340e96d253b4a1..
I almost got lost because it could not keep pace with the speed of play..  

[moocoin]

I'm a little curious as to why my account that went 95-0 didn't win the best win:loss ratio leaderboard.  Seems like 100% should beat 99.68%.

Yep - totally fair question.  We use a weighting algorithm as described in the rules.   More specifically, we use the lower bound of the Wilson interval at a 95% confidence interval to calculate the rankings.  This makes it so that someone who wins 999/1000 games ranks higher than someone who won 1/1 games.  Clearly the guy who won 999/1000 is more likely to be the better player.

Thanks,
MooCoin

That makes sense.  I guess my only argument would be that the leader-board doesn't say "best player" it says "best win:loss ratio".  And a 1:0 ratio is better then a 999:1 ratio.  If you wanted to figure out the best player, it seems like it would be some sort of ELO rating system or something.  It's nitpicky on my end, sure, but the whole leaderboard bot stress test system seemed to be built around being nitpicky.

Nah - I don't think it's nitpicky.  The reality is that we were going to do a straight ratio, then decided to go ahead and do something better (Wilson interval), and neglected to go back and change the subtitle from "ratio" to "ranking".  We did clarify our calculation in the rules but you're right - we should have changed it from ratio.

For the day-to-day play, the leaderboards will be for fun.  Before we run our next tournament (on the MainNet!!!), we're going to implement ELO to take into account the relative strength of players.  This will become even more important as we create more and more skill-based games.

Thanks for the feedback and sorry if it wasn't clear how we were ranking players.
MooCoin

[FUR11]

I'm a little curious as to why my account that went 95-0 didn't win the best win:loss ratio leaderboard.  Seems like 100% should beat 99.68%.

Yep - totally fair question.  We use a weighting algorithm as described in the rules.   More specifically, we use the lower bound of the Wilson interval at a 95% confidence interval to calculate the rankings.  This makes it so that someone who wins 999/1000 games ranks higher than someone who won 1/1 games.  Clearly the guy who won 999/1000 is more likely to be the better player.

Thanks,
MooCoin

That makes sense.  I guess my only argument would be that the leader-board doesn't say "best player" it says "best win:loss ratio".  And a 1:0 ratio is better then a 999:1 ratio.  If you wanted to figure out the best player, it seems like it would be some sort of ELO rating system or something.  It's nitpicky on my end, sure, but the whole leaderboard bot stress test system seemed to be built around being nitpicky.

Nah - I don't think it's nitpicky.  The reality is that we were going to do a straight ratio, then decided to go ahead and do something better (Wilson interval), and neglected to go back and change the subtitle from "ratio" to "ranking".  We did clarify our calculation in the rules but you're right - we should have changed it from ratio.

For the day-to-day play, the leaderboards will be for fun.  Before we run our next tournament (on the MainNet!!!), we're going to implement ELO to take into account the relative strength of players.  This will become even more important as we create more and more skill-based games.

Thanks for the feedback and sorry if it wasn't clear how we were ranking players.
MooCoin

I understand.  Bummer I wasted my time on something that wasn't clear, but it was only a few hours and honestly I'll probably play on the site once it's up for real btc because the hi/low game is really fun and seems like something that has a skill element to it (and I used to love watching Card Sharks).

[moocoin]

I'm a little curious as to why my account that went 95-0 didn't win the best win:loss ratio leaderboard.  Seems like 100% should beat 99.68%.

Yep - totally fair question.  We use a weighting algorithm as described in the rules.   More specifically, we use the lower bound of the Wilson interval at a 95% confidence interval to calculate the rankings.  This makes it so that someone who wins 999/1000 games ranks higher than someone who won 1/1 games.  Clearly the guy who won 999/1000 is more likely to be the better player.

Thanks,
MooCoin

That makes sense.  I guess my only argument would be that the leader-board doesn't say "best player" it says "best win:loss ratio".  And a 1:0 ratio is better then a 999:1 ratio.  If you wanted to figure out the best player, it seems like it would be some sort of ELO rating system or something.  It's nitpicky on my end, sure, but the whole leaderboard bot stress test system seemed to be built around being nitpicky.

Nah - I don't think it's nitpicky.  The reality is that we were going to do a straight ratio, then decided to go ahead and do something better (Wilson interval), and neglected to go back and change the subtitle from "ratio" to "ranking".  We did clarify our calculation in the rules but you're right - we should have changed it from ratio.

For the day-to-day play, the leaderboards will be for fun.  Before we run our next tournament (on the MainNet!!!), we're going to implement ELO to take into account the relative strength of players.  This will become even more important as we create more and more skill-based games.

Thanks for the feedback and sorry if it wasn't clear how we were ranking players.
MooCoin

I understand.  Bummer I wasted my time on something that wasn't clear, but it was only a few hours and honestly I'll probably play on the site once it's up for real btc because the hi/low game is really fun and seems like something that has a skill element to it (and I used to love watching Card Sharks).

Thanks.  We look forward to having you join the live site.  We have even more games in store, as well.

[moocoin]

Just a heads up. 
During the tournament, we collected over 1,100 BTC in TestNet coins.  We just donated these coins back to the TestNet faucets:
550 coins to TP's TestNet Faucet
550 coins to MojoCoin

Here are the transactions in case anyone cares 
http://blockexplorer.com/testnet/tx/cc3c9bf8655479f7afb018907468bda3da2f760ac76fde70a44f5633b501b930
http://blockexplorer.com/testnet/tx/c4dce9e2ac16dae41ab51fe3647501c3fc3d89b0c72809eff0b8b28e151c2004

Thanks,
MooCoin

[RGBKey]

Just a heads up. 
During the tournament, we collected over 1,100 BTC in TestNet coins.  We just donated these coins back to the TestNet faucets:
550 coins to TP's TestNet Faucet
550 coins to MojoCoin

Here are the transactions in case anyone cares 
http://blockexplorer.com/testnet/tx/cc3c9bf8655479f7afb018907468bda3da2f760ac76fde70a44f5633b501b930
http://blockexplorer.com/testnet/tx/c4dce9e2ac16dae41ab51fe3647501c3fc3d89b0c72809eff0b8b28e151c2004

Thanks,
MooCoin

That's very responsible of you. Will there be another contest like this? I missed out on most of this.