Ranking with CGOTM(C3C) and GOTM (PTW)

[Darkness]

A top player sitting out a long time is the same as a new player in terms of global ranking. They would start from scratch. If that player is consistently high scoring, they will move up the ranking, taking up to ten months to move to the top. I was up to 8th I believe (can't check at the moment), and then took 6 months off, and now I'm down around 166th or so. What little is left of my old scores will roll off while I post new ones, meaning my ranking will grow slowly, especially until I get back in the groove.

I agree, Xevious.
After GOTM22 I was ranked 9th, but I have not participated after that, until this month when I played COTM01 as an unranked newbie...

@Ainwood: I also agree with you. Two rankings would be confusing, as well as much more wrk for the staff. I merely offered it as an option (albeit not a very good one). I just think it would be a very bad idea to give either one or the other game a scoring bonus. It would simply unbalance the rankings. The percentage calculated from the top player's Jason score would also be unbalanced if all the top players would play only one game.

Suggestion: The staff could bribe SirPleb to play both games each month until they come up with a better solution (like a completely worked out points system referring to best finish dates, but I think this might take away some of the comparison possibilities between the different victory conditions) ...

 

[dojoboy]

...., they will move up the ranking, taking up to ten months to move to the top.

I guess I should probably go read the FAQ on Global Rankings. But, from what you say, it only considers the last 10 GOTMs?

 

[ainwood]

Even I was a top-10 player, once.

The rankings will depend on a good number of players, with the best players playing in both competitions. And don't get too focused on SirPleb being top-dog, compete for his spot instead. We have some other very good players who are doing just that.

 

[Xevious]

I guess I should probably go read the FAQ on Global Rankings. But, from what you say, it only considers the last 10 GOTMs?

As posted earlier in this thread it's something like difficulty + 5 months to roll off the chart. I believe deity is difficulty 5, so a deity game would take 10 months to drop off, while a regent (difficulty 2) would drop off in 7 months.

 

[AlanH]

.... with the added complication that deity is one higher in C3C to make room for demi-god

EDIT: Correction to my earlier statement. The aging factor is currently based on (difficulty+4) months, not (difficulty+5). Deity is actually difficulty=6 in Classic, difficulty=7 in C3C.

 

[King Of America]

Please design the new system so people are not penalized for only playing 1 game a month.

Perhaps you can design a scoring algorithm that calculates a player's best score, if they submit 2 games (i.e., a COTM and a GOTM classsic), and leave it that.

For lots of good reasons, some of us have time to play 1 competitive game a month and some have time for 2. Please allow both groups to keep on playing in the same arena comfortably (well except for the occasional lack of resources needed for UUs )

 

[ainwood]

Please design the new system so people are not penalized for only playing 1 game a month.

Perhaps you can design a scoring algorithm that calculates a player's best score, if they submit 2 games (i.e., a COTM and a GOTM classsic), and leave it that.

For lots of good reasons, some of us have time to play 1 competitive game a month and some have time for 2. Please allow both groups to keep on playing in the same arena comfortably (well except for the occasional lack of resources needed for UUs )

That is one of the primary goals.

 

[samildanach]

How will the result releases be scheduled? Will the COTM and GoTM results be released at the same time? Or will the COTM results be released during the first half of the month and the GOTM during the second half? Which ever it is I take it we will have to wait until the end of the month for the GPR updates.
Any decision on the scoring process? I assume the jason scores for both the COTM1 and the GOTM32 are unofficial at the moment since you haven't finalised the scoring system. Are you going to let it slide this month and implement the new system next month?

 

[mad-bax]

For me I would have ranking scored as a percentage of Jason best score. I would have no amortisation (older games being worth less). Finally, I would deal with the classic/c3c situation as follows.

Each c3c/classic cycle is divided into slots. Each slot therefore has space for two games.

EVERY game counts, whether it is C3C or Classic.

Only games from the last 8 slots count unless...

There are more than 8 games in the last 8 slots, in which case the last 8 games count.

Where an entire slot is missed (i.e no classic or C3C submission) then the slot is empty and zero rated and the number of qualifying slots or games is reduced by one.

Complicated? Sounds like it, but in practice (or with a picture) it would be straightforward enough IMO.

By this method the parameter that gets screwed with to make a comparison is the time base and not peoples scores.

My 2 cents, and worth rather less.

 

[AlanH]

How will the result releases be scheduled? Will the COTM and GoTM results be released at the same time? Or will the COTM results be released during the first half of the month and the GOTM during the second half? Which ever it is I take it we will have to wait until the end of the month for the GPR updates.

Having raised expectations for prompt results release I don't think we are now going to delay any deliberately. Global Ranking updates will have to wait until both games have been adjudicated, as it's not really appropriate to update the month's rankings first just for C3C players.

Any decision on the scoring process? I assume the jason scores for both the COTM1 and the GOTM32 are unofficial at the moment since you haven't finalised the scoring system. Are you going to let it slide this month and implement the new system next month?

Why? The Jason scores we are issuing for COTM1 are valid, and use the same algorithms as we have set for future C3C games, which are derived from those used for the Classic games. Jason scoring is done on a per-game basis and we'll only be normalising scores for the two games each month to best score for the purposes of Global Ranking. There's no reason to suppose Jason scores are interim or unofficial in any way.

 

[samildanach]

Having raised expectations for prompt results release I don't think we are now going to delay any deliberately. Global Ranking updates will have to wait until both games have been adjudicated, as it's not really appropriate to update the month's rankings first just for C3C players.
Why? The Jason scores we are issuing for COTM1 are valid, and use the same algorithms as we have set for future C3C games, which are derived from those used for the Classic games. Jason scoring is done on a per-game basis and we'll only be normalising scores for the two games each month to best score for the purposes of Global Ranking. There's no reason to suppose Jason scores are interim or unofficial in any way.

Thanks. Perhaps I misread it but I thought the prefered option, at the moment, was to normalise against best dates as opposed to best player. The reason ainwood gave for this was that the classic games might not attract as many elite players - therefore the scores of the player pool in classic games might recieve a boost as their scores would potentially be normalised against a lower best score. The reason I assumed that the results were not official as yet as I thought the best dates for the victory conditions were having to be calculated with a greater degree of stringency to ensure parity between the two competitions.

 

[ainwood]

If we find that the player pool for either of the games is dropping to the point where the relative scoring seems unfair, then we will consider modifying the scoring. We could perhaps do a 'reasonableness' test by comparing how players competing in both competitions are faring in similar games.

 

[tao]

Each c3c/classic cycle is divided into slots. Each slot therefore has space for two games.

EVERY game counts, whether it is C3C or Classic.

Only games from the last 8 slots count unless...

There are more than 8 games in the last 8 slots, in which case the last 8 games count.

This seems like a very good approach IMHO.

Where an entire slot is missed (i.e no classic or C3C submission) then the slot is empty and zero rated and the number of qualifying slots or games is reduced by one.

I'm not so happy with this, because it slightly disfavors people (like me ) that cannot play both versions. But overall this scheme looks good.

 

[civ_steve]

The 'slots' approach doesn't allow higher difficulty games to persist longer than lower difficulty games. That is an important element of the current scoring system, IMO.

In fact, that is a tricky feature to implement if 2 monthly submittals are combined, either direct or weighted average.

 

[scoutsout]

This may sound a little off-the-wall... but have you guys ever discussed a "handicapping" system analagous to golf? I offer this suggestion as an adjunct, not an alternative, to the current global ranking system.

As I understand it, in order for a golfer to have a handicap, nine scorecards must be turned in. In computing the handicap, a "slope" rating is used (analagous to Jason scoring) that attempts to equalize courses that vary in difficulty. I could be wrong, but I believe the handicapping averages the best 9 of the last 18...

...so this might be another thing that would encourage continued participation... if you have to play 9 games to even GET a handicap....

Though this wouldn't really use "all" of the GsOTM that a player submits, games would age more slowly...

Edit: corrected a typo, and had this thought: Handicapping would measure a players performance in a given against their "expected best" performance. By averaging the best 9 of the last 18, and comparing that to a given game, they would have pretty good feedback on whether their last game was up to their own standards...

...or something like that.

Bear in mind, such a system would be more useful for rating a player's performance relative to their "expected" performance (rather than rating someone like me against someone like DaveMcW or SirPleb) but it might give some of those in the "middle of the pack" something else to shoot for... "beating our handicap".

 

[akots]

If we find that the player pool for either of the games is dropping to the point where the relative scoring seems unfair, then we will consider modifying the scoring. ...

It would be nice to know this in advance but not after the particular GOTM was played. For example, this month I was unable to play COTM1 and would have hard time trying to play GOTM32 due to RL issues. Next month, I would like to go for both, if RL permits. In this case, there is no sesne to try GOTM33 if it would be disregarded or scoring modified to give some "fair" result.

 

[bradleyfeanor]

So do we have a final verdict on how ranking will be handled with COTM/GOTM? The time draws nigh...

I am only curious because it may affect whether I play GOTM 33 or not.

 

[Sandman2003]

So do we have a final verdict on how ranking will be handled with COTM/GOTM? The time draws nigh...

A very good point. Do we have a decision as to how this works? Personally I favour a best of the two scores approach as the best way to not disadvantage thoise who only play one, while encouraging some players who can play both to play variations. The only issue I see is how to you work the situation where the difficulty level in one is not the same as the other eg Cotm uses monarchy, Gotm uses emperor in one month. Some players will have their better score in cotm, but this score will persist for one month less than an equivalent Gotm score?

 

[Sir Bugsy]

Personally, I don't like the aging process, (either here or in RL ). Presently, I'm ranked 26th. There is NO WAY, I'm better than someone like Dave McW, but according to the rankings, I am. Why? He has played once in the last 10 months (and won BTW) and I have played 8 times. If you count up all his victories and medals (which happen to be more than all the GOTMs I've played) Dave (and Sir Pleb) would be at the top of the list.

How to solve it? I'm not sure, but I do know that the rankings don't reflect reality.

And please don't take this as a slam on the GOTM Staff. They do a fabulous job. I just don't think the present system is working.

 

[samildanach]

I'm not sure, but I do know that the rankings don't reflect reality.

The reflect reality, in as much, as you have played more than they have.
There may be all sorts of reasons why a player does not submit. They may have had a bad game and decided to abandon it. Grown board with civ. Or perhaps the real world has intruded for a while and they've not been able to play. I think that in the latter case it is unfortunate given the amount of time a player must spend to climb up the rankings.
Since AlanH wasn't able to play this month because his staff duties conflicted with his ability to compete in the GOTM I expect to pass him in the rankings. He is probably a better player than I am but does that bother me......no because I will get to beat my chest and post dancing banana cartoons Lots and lots of dancing bananas.