No one wants a cookie. They want to know if you're going to bail on a game because it's time for your period again. No one chooses a game over life. No one. Some of us manage our gaming commitments around life without inconveniencing others. Life does happen. It happens to everyone. Some people just manage it with decency and respect for others. This "real life" excuse is just another crutch for a cast of losers to present themselves are reliable players. Isn't it interesting that "real life" never happens to someone who is winning a game, only to players who just got stabbed...

The conversation isn't pointless. If there is going to be a RR, it should mean something The players signing up for games have a right to know whether their fellow players are going to just walk out on a game (again).

Imagonnalose has 134 missed phases out of 2634. That's 95%. Give this guy a cookie, someone, please. He's ONLY missed 5% of his moves. That's damn near one missed phase per game/ I guess real life happens a lot to some people. But hey, you picked your nickname well, if that's how you're are going to play.

My point of view is different, Imagonnalose:

if you're not able to follow a particular game, you don't join it, or you opt for a longer phases one.
This site does not lack variety of the offer.

So I'm hearing that a heavier weighting for missed phases in the RR might be a good thing?

I guess the point of the conversation is what exactly is the RR for. If we are going to have one at all, it may as well measure something. If there's a concern about making the slackers feel like someone stole their cookies, then just do away with it altogether.

I liked Tomahaha's suggestions.

Immagonalose:
It would take at least half an hour to check even a few games for every player before starting a game. It would take even longer with a larger game. It also says nothing for anonymous games in which RR is used to keep the game limited to high RR players. This way players that want to commit can play with each other with little fear of a CD, while players that have CDed can play together and work their way up to that point. Without that, there is too much unpredictability.

I think from looking at Imagonnalose's "What's This" next to his reliability, it gives a clue as to the issue. The missed phase calculation seems good enough, but then his CD - Taken over CD penalty is: 10 * (21 - 21) = 0

So basically he is getting 0 penalty against his reliability for 21 CDs (!?!). CDs should have more of an impact on reliability than NMRs, and NMRs have a 2x multiplier. So maybe CDs have a 4x multiplier.

So based solely on his phases, his reliability is 89.83. His 21 CDs currently have no impact on this number. Implementing a 4x multiplier of CDs against games played, we have 100 - (100 * (4 * 21) / 236) = 64.4

So some way to combine phase reliability and CD reliability would be good, or just display them both.

I like that idea Mapu, using this example what would the RR look like with the 2x multiplier for NMRs in this equation?

So for me, I've CDd in 3 games, and have missed 55 phases out of 5362 played.
something like this?

100 - (100 * (2 * 55) / 5362)) - (100 * (4 * 3) / 5362)

The 2nd 5362 would be your number of games played, so 317. Your reliability would be 94.16.

The good part of my formula is that you can improve your percentages over time by being reliable.

That's a good question Mapu.

@Oli, did you have to add tables/columns to track missed phases or was it possible to calculate with queries based on the ones from webdip?

Is there a way to include an incentive for players to take over positions in CD?

For instance, in this equation there is no accounting for active games, only games finished. Some sort of ameliorating section that takes into account games active to provide incentive for players to slightly improve their RR by taking over positions in CD?

Yes that's a great point. Perhaps a 3rd equation that added a pct back on without masking the CDs like in the current formula.

The first part is: 100 - (100 * (2 * NMRs) / Phases Played)) - (100 * (4 * CDs) / Games Played)

The only problem is you don't really prove how reliable you are by taking over CDs. That's why it's hard to formulate.

So again using me as an example, let's pretend I'm currently playing in 5 games.
something like this?

100 - (100 * (2 * 55) / 5362) - (100 * (4 * 3) / 317) + (100*(6 * 5) / 5362)

making an RR of 94.71

"The only problem is you don't really prove how reliable you are by taking over CDs. That's why it's hard to formulate."

^^This

Taking over a CD does NOTHING to prove you are reliable. The only way to prove you're reliable is to play more games WITHOUT missing phases and going CD. Taking over CDs should have nothing to do with how reliable a person is.

CDs used to be a real and significant problem before the RR system was implemented. It would be a shame to remove the incentive for players to take over countries in CD.

100 - (100 * (2 * NMRs) / Phases Played)) - (100 * (4 * CDs) / Games Played) + (100 * (6 * Games Active) / Phases Played)

Can you figure out Thrawns using that formula as he was another good test case? I do like the active games part.

GA Thrawn:
100 - (100 * (2 * 33) / 1774) - (100 * (4 * 5) / 80) + (100 * (6 * 6) / 1774) = 73.308

Imagonnalose:
100 - (100 * (2 * 134) / 2634) - (100 * (4 * 21) / 80) + (100 * (6 * 3) / 2634) = -14

So something is flawed...

Ima has 236 games played, not 80.

54.92 for Ima

And then someone like RUFFHAUS who has missed one phase is: 99.946 reliability

I think there needs to be a cap where your active game bonus cannot exceed your CD penalty.

So I think it works.

My point is simply this: you are complaining about something someone else created. RR is a tool that was supposed to give a very rough estimate of a players quality.

All in all, my primary point is that YOU have the control over whether or not to join a game. Do not blame someone else because you decided to join a game with someone who has missed a phase. That's why I play with only a few people and even with those, only in certain situations. Stop being all self righteous because you feel better about yourself because the most important thing in your life is this game.

If I join an anon game where you are entered at 90+ instead of 54 where you should be how could I know you are so unreliable?

I wonder if it would be possible to have the RRank displayed for a little while alongside the current RRating system. And then ask people what they prefer?

A recap so far:

RRank:
Using this formula some examples given so far are:
100 - (100 * (2 * NMRs) / Phases Played)) - (100 * (4 * CDs) / Games Played) + (100 * (6 * Games Active) / Phases Played)

GA Thrawn:
100 - (100 * (2 * 33) / 1774) - (100 * (4 * 5) / 80) + (100 * (6 * 6) / 1774) = 73.308
Imagonnalose:
100 - (100 * (2 * 134) / 2634) - (100 * (4 * 21) / 263) + (100 * (6 * 3) / 2634) = 54.92
Kaner406
100 - (100 * (2 * 55) / 5362) - (100 * (4 * 3) / 317) + (100*(6 * 5) / 5362) = 94.71
Ruffhaus 8
100 - (100 * (2 * 1) / 3726) - (100 * (4 * 0) /72) + (100*(6 * 4) / 3726) = 99.46

****
Stats:
Display a record of CDs
Display a record of how many CDs a player has taken over

****
Suggestions:
weighting CDs in recent games more heavier than older CDs.
display a record of CDs leading to being defeated (if possible)
display the RRank for preliminary discussion

****

Questions
1. did you have to add tables/columns to track missed phases or was it possible to calculate with queries based on the ones from webdip?

One small correction is that RUFF would not have the 3rd active game bonus because he has no CDs. If you kept it in, it would be possible for people to go over 100 (like I would for example).

Also, if there could be a reliability rank like a points rank (ie RUFFHAUS : 99.946 (6 of 1000) or whatever) with a tiebreak of phases completed.