A proposal for a slight modification of Jason scoring

[solenoozerec]

here

It was fast. and

First, I think, it makes sense. There are still the same people at the top, they exact positions are slightly altered. This tells me that both systems (Jason score and best date based) work correctly.

As for which one (Jason dates or fastest player) is better – I do not know. IMO fastest player is better, because it underlines and enhances the difference with current Jason Score system. In case of fastest player, score depends less on the game itself and more on competition.

 

[AlanH]

I think overall positions in the table are more to do with regular play then skill. That's also true of the GPR table to some extent, but I need to separate the scores more as it's emphasised more here. You should probably focus on whether the per-game results look reasonable.

I don't think we need to avoid using Jason dates just because ... The Jason dates have relatively little impact on the existing scoring system, so why not make more use if them here? What's needed is the best way to identify differences in speed of victory across the conditions. If that requires independent target dates rather than the fastest players on the day, so be it.

 

[klarius]

a bit arbitrary - why not a cubic or quartic?

Sure why not.
Applying any exponent is just the easiest way to map a 0..1 range on itself with a changed distribution.
The best exponent could be found by trying it.

I suggested squared just because I was a thereotical physicist in a former life and by that cannot count further than 2.

Edit:
A suggestion for a less arbitrary exponent would be:
Find the median of the percentages.
Calculate the exponent as log(0.5)/log(median).
That will map the median to 0.5 .

 

[Offa]

I admit I haven't followed all this, partly because I don't mind the Jason score too much. Sure it favours domination/conquest but so what: it's better than the old system that almost necessitated milking.

I am intrigued though about ideas to change the global ranking. At present this is largely a reward for perseverence and I would support any sensible effort to make it reflect the quality of play instead. At present a player with a gold medal 100% one month followed by a no show the following month, would score a bit less than a 50% effort in both months. There are all sorts of ways this could be changed. I rather support the idea that every game, however bad, should increase your ranking, but the weightings for top games need to be beefed up. For example Drazek hasn't played for a month and suddely he's down to tenth. Perhaps fairly poor Offa, admittedly not ranked that high in the first place , hasn't played for a month and has almost vanished off the chart.

As it is, I hope that most players don't worry too much about the scores. I think it is possible to get a pretty good feel for how good players are just from the posts, and anyway, it's all supposed to be a bit fun.

I think a bonus for players with short usernames should be considered, especially those derived from 8th century monarchs.

 

[solenoozerec]

I think a bonus for players with short usernames should be considered, especially those derived from 8th century monarchs.

:shakehead:

On topic: I think the way GR is designed is to reward regularity in play.
If you want your name to be memorized, go for medals and awards.

 

[delmar]

I don't think we need to avoid using Jason dates just because ...

Thanks for putting that web page together so quickly.

I would intuitively vote for the actual best dates because I believe that would encourage a more equal spread of talent across victory conditions. Thanks to those "philosophical" discussions, I realized that I consider this the most important objective here because it would allow people to compare themselves to one of the best players regardless which victory condition they prefer.

As a side note, my impression is that the Jason date is realistic or even too high compared to the actual achieved conquest and domination dates and is a bit too agressive for the other victory types. I think some people aluded to this earlier but I don't know if there was ever a comparison. Was there? If not, would it be possible to do a comparison for the past GOTMs? Note that I am not asking whether the Jason dates could have been achieved for a given condition if only a top player had chosen to go for it. I am curious how the actual best dates compared to the Jason dates for a given game.

 

[AlanH]

I think a bonus for players with short usernames should be considered, especially those derived from 8th century monarchs.

Global rankings are, and should be, rewards for peristence. We have rewards for per-game performance in the game rankings and awards. We have recognition for life-time achievements in the Pantheon. The Global Rankings are intended to show who's hot over the last few months, and who's not. Just like season stats in other games. I'm just trying to see if there's another axis to look along to see who's hot, that would look at date alone instead of emphasising expansion.

Having watched the debate rage, here's my thinking on the way military victories dominate the per-game results. It's probably been said here already in a hundred different ways, but treat this as my summary and thinking aloud:

Fast domination dates are achieved by getting to 67% territory & population as fast as possible - typically before the middle to end of the Middle Ages. To achieve a high Jason score with a domination vctory you monitor happy face count in addition to territory. But as happy faces are largely proportional to population in the early game, which is largely proportional to city count, which mostly comes back to territory, this hardly requires any compromises. A fast domination victory will produce a high Jason score with very little balancing required. Conquest happens early as well, and it usually happens over a short period in the end-game when keeping or disposing of captured cities and territory will not change your score average much. There may be a small balancing effect, but well-played conquest games are going to produce very similar scores whether territory is kept or not.

For any other victory condition, territory and population are means to the end, if the end objective is a fast victory. So targeting a high Jason score for these conditions *must* involve significant effort to find and maintain a balance between expansion and speed.

As a bear with very little brain, and a guy, I can only walk or chew gum. Trying both would do permanent damage to my brain cell and my bent chromosome. So it's not surprising that I choose domination as my default victory condition. Anything else would involve trying to balance and compromise expansion vs speed in a situation where I know the expansion objective but I don't know how fast is fast enough. I may not be typical, but I suspect many players share my simple approach to choosing a victory condition. And once you find a groove you tend to stick in it.

 

[AlanH]

I am curious how the actual best dates compared to the Jason dates for a given game.

I'll put together a database report comparing fastest date against Jason date by game and condition. It's been on my to do list since the debate started, but tonight it comes below the QSC results. Watch this space.

 

[AlanH]

it would allow people to compare themselves to one of the best players regardless which victory condition they prefer.

I understand where you're coming from, but I've looked at the numbers of people playing different victory conditions. When you get very few players going for culture or diplo in a game you really don't have a valid basis for comparative scoring against the results of 10 or 20 times more domination or conquest players. As Aeson says, a middle-of-the-road player's score in the date-based table will depend almost totally on good or bad luck. Did SirPleb choose this victory condition this month?

I'd love to believe that market forces would prevail, but this isn't a big market. Fewer players will enter 400 turn games than 200 turns - fact of real life. We'll never have more than maybe 10 players per late victory condition unless we set a Sponsored Victory a la Medal Play.

 

[samildanach]

I admit I haven't followed all this, partly because I don't mind the Jason score too much. Sure it favours domination/conquest but so what: it's better than the old system that almost necessitated milking.

I am intrigued though about ideas to change the global ranking. At present this is largely a reward for perseverence and I would support any sensible effort to make it reflect the quality of play instead. At present a player with a gold medal 100% one month followed by a no show the following month, would score a bit less than a 50% effort in both months. There are all sorts of ways this could be changed. I rather support the idea that every game, however bad, should increase your ranking, but the weightings for top games need to be beefed up. For example Drazek hasn't played for a month and suddely he's down to tenth. Perhaps fairly poor Offa, admittedly not ranked that high in the first place , hasn't played for a month and has almost vanished off the chart.

Given that there are a hat full of players now with GPRs above 400 it was hardly suprising. And remember Drazek got to be No.1 by being the most consistent player IIRC he didn't win any gold medals: he stopped being consistent, he stopped being No.1. It's as simple as that.

The same thing goes on further down the ranking Qwertysoft missed a deadline on a game which he hoped would take him onto the first page and ended up on the third page.

As it is, I hope that most players don't worry too much about the scores. I think it is possible to get a pretty good feel for how good players are just from the posts, and anyway, it's all supposed to be a bit fun.

Its funny you should say that but there are many players in the SGs forum (I don't mean the SGOTM run by M.B. ) who talk a very good game. But have never played a GOTM or COTM. There are of course excellent, former stalwarts of the GOTM who hang out there but what about the others?
[Football coach mode]Keeping your focus for five or 10 turns every few days is a bit different from playing 200-540 turns on your own. Civ like chess is not just about how much you know the theory its also about stamina, psychological toughness and focus[/Football coach mode]

I personally dislike people who set themselves up as gurus but do not play in competition even though they have opportuinity to do so. On the other hand I read everybodys posts in the spoilers just the same whether its an old or a new player for something that might improve my game in future. Big respect to everyone who posts and/or plays in the COTM/GOTM. No respect to self-appointed gurus.

 

[CKS]

Thanks, AlanH, for putting together this table. As I moved up about 30 places (and onto the first page!), it made my afternoon a happy one.

I don't tend to expand very much, and I like 20 K and space ship victories, so I don't head down the conquer-the-world path. I rarely start wars with the AI (it seems wrong to invade them). Between my lack of skills (which I'm improving) and my dislike of invasion (which won't change much), I'll probably rarely score very well using either Firaxis or Jason. I like having another way to compare my game.

Also, thanks for the consolation prize for losing - that's about half my games so far. What I'd suggest as a slight change would be giving longer surviving losers more points than sooner dying ones: turns survived/100 (or some other number, this gives 5.4 points maximum, 180 would give 3 points max). Lasting longer while still losing means I'm doing better.

 

[AlanH]

Yes, the loser score variation has been mentioned as a refinement.

 

[Sandman2003]

And I thought I would go up to, but I actually slip 3 places. I suspect that there is a huge boost to some players who regularly have submitted in the less popular VCs, and rgar regular participation has become even more valuable as AlanH suggested above.

The spread over max-min dates sounds like a good idea!

 

[denyd]

I kind of like this alternative chart. I went from 36th to 18th (only 4 spots behind SirPleib).

This does kind of illustrate who a persistant player (I have entries for each set) can reach the top-20 through sheer participation.

I'm am a little surprised that I went up so far. Most of my recent month's entries (at least the last 5) have been spaceship victories, which by definition are not a quick as domination / conquest type wins.

 

[delmar]

I understand where you're coming from, but I've looked at the numbers of people playing different victory conditions. When you get very few players going for culture or diplo in a game you really don't have a valid basis for comparative scoring against the results of 10 or 20 times more domination or conquest players. As Aeson says, a middle-of-the-road player's score in the date-based table will depend almost totally on good or bad luck. Did SirPleb choose this victory condition this month?

SirPleb is surely exceptional but he is not the only great player. So it's not like the choices of 1 person will swing the score from one end of the spectrum to the other. I think if we can get any one person from the top ten to choose a particular victory condition then that particular victory is already better off than with an estimated Jason-date. Will we get one top player for all 6 conditions? I don't know. But it can't get much worse than right now.

I'd love to believe that market forces would prevail, but this isn't a big market. Fewer players will enter 400 turn games than 200 turns - fact of real life.

Significantly fewer? Even if they can score a lot higher? Is this statement true if we examine the population of top players (who are presumably more enthusiastic than the average)? 'Cause we don't need all 100+ people be evenly distributed, only the top dogs...

More importantly, is this just your gut feel or was something like this attempted earlier?

We'll never have more than maybe 10 players per late victory condition unless we set a Sponsored Victory a la Medal Play.

What exactly is a "Sponsored Victory a la Medal Play"? Is it possible perhaps that basing the score on the best actual date would make all the less popular victories automatically "sponsored"?

 

[Offa]

Its funny you should say that but there are many players in the SGs forum (I don't mean the SGOTM run by M.B. ) who talk a very good game. But have never played a GOTM or COTM. There are of course excellent, former stalwarts of the GOTM who hang out there but what about the others?

I personally dislike people who set themselves up as gurus but do not play in competition even though they have opportuinity to do so.

What if those guys did play on GOTM? They might actually play really well and make our scores look even worse. Like you, I am a big fan of GOTM, and probably wouldn't play at all otherwise (apart from a few tiny map conquests if I can ever get mapfinder to work again).

However, I think you can tell the skill level of new players pretty quickly from their write up of a GOTM. I fear that this makes me sound a bit like a teacher marking coursework, but really I am just keen for players to write about a game which I have played myself. At the moment there is a rather odd tendency for there to be many more posts in the pre game threads (not to mention this thread) than in the spoilers. Do people really have more to say about a game before they start it, than after they have played it?

 

[AlanH]

SirPleb is surely exceptional but he is not the only great player.

No. He's one example of the kind of player who could drastically affect the fastest date in any one game/condition.

So it's not like the choices of 1 person will swing the score from one end of the spectrum to the other. I think if we can get any one person from the top ten to choose a particular victory condition then that particular victory is already better off than with an estimated Jason-date. Will we get one top player for all 6 conditions? I don't know. But it can't get much worse than right now.

It's a small group. I wouldn't bet on them spreading themselves around.

Significantly fewer? Even if they can score a lot higher? Is this statement true if we examine the population of top players (who are presumably more enthusiastic than the average)? 'Cause we don't need all 100+ people be evenly distributed, only the top dogs...

The smaller the group you want to spread evenly the less likely it is to happen.

More importantly, is this just your gut feel or was something like this attempted earlier?

Statistical reality. Tell me how many players you would trust to deliver benchmark performance across all victory conditions in all game situations. 10? 20? Then work out how they will spread over 7 victory conditions, remembering they won't make random choices and they may not all play every game. Depending on what they can see of map, difficulty, civ, start position and early game development they'll determine the "best" condition for the map. They won't all make the same choice, but they are unlikely to choose all 7 options between them. Even if they chose randomly you'd be surprised how often statistically you would get none of them choosing one of the conditions.

What exactly is a "Sponsored Victory a la Medal Play"?

We had a few Medal Play games where, as well as counting for normal GOTM scoring they could be played to a specified victory condition for a straightforward fastest finish medal.

Is it possible perhaps that basing the score on the best actual date would make all the less popular victories automatically "sponsored"?

I don't think so. The idea was to provide an extra incentive for more players to pursue a common victory condition in a particular game.

 

[solenoozerec]

I kind of like this alternative chart. I went from 36th to 18th (only 4 spots behind SirPleib).

I went down by 12 positions, but I also like it. 100 sounds more solid than 88 (though 88th is amazingly symmetrical).

In your case, I guess, you went up because you do not play as aggressively as you have to when you compete for Jason score.

Do people really have more to say about a game before they start it, than after they have played it?

In my particular case, when pre-game discussion is going I do not play any games, but I have a temptation towards new one and I also have time to visit this forum.
When I play a game, I have a choice between writing a spoiler and playing. I make probably wrong, but natural choice. When game is finished, I do not remember how this game was started.

 

[AlanH]

In your case, I guess, you went up because you do not play as aggressively as you have to when you compete for Jason score.

More likely because you play consistently and some of the people above you in the real GR table don't. This table puts much more emphasis on regular play, because of the compressed points range in each game. I recommend you focus on how players rank in each game, not on the total in the left hand column.

 

[akots]

This date bonus might of interest only provided that it is given for a certain only one particular victory condition in a given GOTM like it has been for the Medal Play series. Otherwise it would just result in further inflation of the ratings and create additional chaos in attempts of trying to guess the score. Otherwise, having all victory conditions count makes little sense since there should be only one ultimate score in the end.

But this is a rather personal opinion which is about adding apples to oranges getting strawberries in the end.

There are also a number of tricks/exploits including Palace jump (and Palace move; in C3C as well not only in PTW), RoP abuse, ship chaining/hopping, etc as it has been discussed many times, which dramatically alter the finish date. However, if the game is well-maintained in therms of happiness, a player can still count on a good score without using these exploits. Adding the finish date bonus would nullify these attempts by making them futile and would directly encourage their use which can now be avoided still getting more or less suitable score. This is a not so personal opinion IMO, just the way of how game mechanics work.