Which game did you download?
HOF Questions and Answers Thread
- Thread starter Methos
- Start date
babaBrian
Rock Chalk!
Vanilla: (didn't work)
19482
19484
BTS: (worked)
19496
Thank you for your help!
19482
19484
BTS: (worked)
19496
Thank you for your help!
Misotu
King
- Joined
- Jan 10, 2006
- Messages
- 934
Hallo, sorry if this has already come up somewhere else but if it has, I couldn't see it. There seems to be a problem with EQM scoring - not sure if it's to do with the (very nice 
) new-look tables. One example: my two games for Mansa (Relig, Large, Deity, Quick) score 10 and 0.3 (!) respectively, but the ad hoc score query shows 59.81 and 23.08 ... not sure why this is?
) new-look tables. One example: my two games for Mansa (Relig, Large, Deity, Quick) score 10 and 0.3 (!) respectively, but the ad hoc score query shows 59.81 and 23.08 ... not sure why this is?
I am thinking that the player of those two games doesn't have BTS. BTS patched Vanilla, changing the DLL. The only way to open the save would be to get the original DLL file from 1.74 and swap it out.
We don't require players to download the new vanilla DLL for HOF since the mod overrides the DLL with our's. (GOTM does for obvious reasons.)
The new page doesn't do any of the calcs, it just displays them. Where are you seeing the 10 & 0.3 scores? Can you send a screen print.Hallo, sorry if this has already come up somewhere else but if it has, I couldn't see it. There seems to be a problem with EQM scoring - not sure if it's to do with the (very nice
Dumb question, but one I do not have an answer to....
Can you submit GOTM games for the HOF ??
Can you submit GOTM games for the HOF ??
- Joined
- Mar 26, 2007
- Messages
- 7,920
No. Games submitted for HOF must have the Game Start generated by the Player himself (including the use of MapFinder) and no Game starts can be the same as any other; nor can they be shared with other players.
All competitors of a GOTM competition start with the same particular instance of a Map (technically map sharing under HOF rules) and that is not permitted in the HOF.
EDIT: See http://hof.civfanatics.net/civ4/rules.php?show=disallowed
Disclaimer: I'm only a HOF Player (not HOF staff), but I'm absolutely sure the above statements are accurate.
Sun Tzu Wu
Thanks Sun Tzu Wu
I suspected as much, but I thought I might as well ask.
I suspected as much, but I thought I might as well ask.
Just confirming that Sun Tzu Wu is right, no XOTM maps allowed for HOF.
Misotu
King
- Joined
- Jan 10, 2006
- Messages
- 934
My Mansa score for the two games combined is shown as 10.3 on the Deity EQM League of Nations screen, which anyone can view so probably no need for a screen print? Clicking on it shows the separate scores of 10 and 0.3 for the two games ...
Kwibus
Chieftain
@ Misotu
Ik think the scores for your games for the EQM are lower because for the EQM the inca and non ancient start games are removed.
In your case the average finished date became earlier when those games were removed.
Your games thus score lower because the average finish date for eqm vallid games is earlier than the average finish date of all the games in that HOF table.
Ik think the scores for your games for the EQM are lower because for the EQM the inca and non ancient start games are removed.
In your case the average finished date became earlier when those games were removed.
Your games thus score lower because the average finish date for eqm vallid games is earlier than the average finish date of all the games in that HOF table.
Deity EQM was what I needed to find it. I have been looking at this but haven't been able to see anything obvious. The game_info on those games doesn't show any date score for me. Could you screen print the 10 and 0.3 you see?
There aren't any inca games on the table in question. So it's not that.
I think it because Sun Tzu Wu has 5 games in the table and Misotu has 2 and that's it. The Q-Score Query page counts all 5 of Sun Tzu Wu's games but in reality only the top 2 count, which are all below Misotu's first entry dropping the score down.
Maybe
Maybe
That's very likely. Since only one game for each player is used in calculating the QScore. Losing first place appears to have cost a lot of points. Only more players hitting the table can smooth out the scores.
Misotu
King
- Joined
- Jan 10, 2006
- Messages
- 934
Thanks for the replies and sorry about the delay in doing these screens but I'm not that great at this sort of thing. OK, I probably haven't been paying attention or something 
or maybe I'm doing something blindingly stupid 
but I don't understand why the QScore query shows this for the EQM score:
but the deity League of Nations league table screen shows this:
and if you click on the number 2 in my Mansa box on the deity league of nations screen pictured above you see these two games:
but there's no date information there to show an EQM score
However, If I view them in my EQM Deity "status" screen, it looks like this:
which is where I can see 10.0 and 0.3 for the 2 games respectively.
Also, I wonder why all the games in the ad hoc query screen are shown as scoring 0.0 for QM as opposed to EQM?
If Ozbenno is right about the reasons why there is a disparity between what the ad hoc query shows and what is shown in the league tables, then I think this a bit bizarre since surely they should perform the same calculation? Er ... but then why does this QScore Query, in similar circumstances:
come up correctly as follows in my EQM Deity "status" screen ...
I think there's something slightly wrong here, honest.
As an aside, I thought the problem with the best game showing up as 100 while the second best gets a seriously low score had been solved by changing the way the calculation was done? I think something about a fake long date was mentioned, which would smooth out the curve. Otherwise, there's not much point in trying for the very hard games where competition is sparse ...
(Not that I'm suggesting my Mansa games were particularly hard you understand 
)
I know you already do loads and I'm sorry to be whinging
or maybe I'm doing something blindingly stupid but I don't understand why the QScore query shows this for the EQM score:but the deity League of Nations league table screen shows this:
and if you click on the number 2 in my Mansa box on the deity league of nations screen pictured above you see these two games:
but there's no date information there to show an EQM score
However, If I view them in my EQM Deity "status" screen, it looks like this:
which is where I can see 10.0 and 0.3 for the 2 games respectively.
Also, I wonder why all the games in the ad hoc query screen are shown as scoring 0.0 for QM as opposed to EQM?
If Ozbenno is right about the reasons why there is a disparity between what the ad hoc query shows and what is shown in the league tables, then I think this a bit bizarre
since surely they should perform the same calculation? Er ... but then why does this QScore Query, in similar circumstances:come up correctly as follows in my EQM Deity "status" screen ...
I think there's something slightly wrong here, honest.
As an aside, I thought the problem with the best game showing up as 100 while the second best gets a seriously low score had been solved by changing the way the calculation was done? I think something about a fake long date was mentioned, which would smooth out the curve. Otherwise, there's not much point in trying for the very hard games where competition is sparse ...
(Not that I'm suggesting my Mansa games were particularly hard you understand ) I know you already do loads and I'm sorry to be whinging
If Ozbenno is right about the reasons why there is a disparity between what the ad hoc query shows and what is shown in the league tables, then I think this a bit bizarre
...
come up correctly as follows in my EQM Deity "status" screen ...
As there are no more than 2 entries per person in that second table, I think it shows that that is the reason for the problem in the other table. As to why the ad hoc query and Q-score show differently
Misotu
King
- Joined
- Jan 10, 2006
- Messages
- 934
It's probably the case that the number of entries is the problem as you say. But since I'm second in both cases, it really shouldn't affect the score like this for one game but not the other, should it?
I'm not bringing this up because of this one game BTW - there were quite a few anomalies around the place last time I looked and I'm sure I could find loads if I tried. Some scores are too high, some too low ...
I suspect it's attributable to this thing about only the best game for each player counting for the QScore. I'm not really sure why counting multiple entries is a problem to be honest, unless the concern is that people will artificially inflate their scores this way. But I can't see that happening really. So how does it work when people enter multiple games with the same settings but different leaders to get the League of Nations section completed? Does each game count because the settings are unique or ... ?
I'm not bringing this up because of this one game BTW - there were quite a few anomalies around the place last time I looked and I'm sure I could find loads if I tried. Some scores are too high, some too low ...
I suspect it's attributable to this thing about only the best game for each player counting for the QScore. I'm not really sure why counting multiple entries is a problem to be honest, unless the concern is that people will artificially inflate their scores this way. But I can't see that happening really. So how does it work when people enter multiple games with the same settings but different leaders to get the League of Nations section completed? Does each game count because the settings are unique or ... ?
Misotu
King
- Joined
- Jan 10, 2006
- Messages
- 934
OK, in the interest of looking like a good guy I went and did a quick trawl and here's one I found where I benefit from the scoring anomaly. And this is definitely caused by the "only one game counts" rule, which I am really beginning to think is causing a lot more problems than it solves. Two time games, identical settings, note the scores:
Two games, one better than the other. They *both* get 100 for QM. They both get zero for EQM of course, because of the non-ancient start. But trust me, they count for 100 each in EQM Deity Rock of Ages, Industrial start.
Really, this isn't any good because in fact there are a lot of games out there getting much higher scores than they should. WastinTime and I are particular beneficiaries from what I've seen from a casual look because we often submit more than one game with the same settings.
So, because only 1 game counts for each player in each category, you can submit multiple games, all of which will score 100 provided you are the leader and no-one else submits a game that beats yours because the only person you are competing against is yourself - and the system deliberately ignores your other games! This is really quite distorting. The only straightforward solution I can see (not being technical) is that every game submitted needs to count for the qscore (except Inca/non-Ancient as now of course).
I bet I've got quite a few games scoring 100 that shouldn't
I went and did a quick trawl and here's one I found where I benefit from the scoring anomaly. And this is definitely caused by the "only one game counts" rule, which I am really beginning to think is causing a lot more problems than it solves. Two time games, identical settings, note the scores:Two games, one better than the other. They *both* get 100 for QM. They both get zero for EQM of course, because of the non-ancient start. But trust me, they count for 100 each in EQM Deity Rock of Ages, Industrial start.
Really, this isn't any good because in fact there are a lot of games out there getting much higher scores than they should. WastinTime and I are particular beneficiaries from what I've seen from a casual look because we often submit more than one game with the same settings.
So, because only 1 game counts for each player in each category, you can submit multiple games, all of which will score 100 provided you are the leader and no-one else submits a game that beats yours because the only person you are competing against is yourself - and the system deliberately ignores your other games! This is really quite distorting. The only straightforward solution I can see (not being technical) is that every game submitted needs to count for the qscore (except Inca/non-Ancient as now of course).
I bet I've got quite a few games scoring 100 that shouldn't
In looking into this a little more, I have noticed that every QScore Query shows a maximum of 100 QM points for each game (even Duel settler games), which isn't right so unless you're talking about Huge Deity games with Raging Barbarians, don't rely on that.
We'll have a look into the scoring glitches.
We'll have a look into the scoring glitches.
The #1 entry always gets a base Qscore of 100 points. The remaining games are scored based on the minimum date and average dates of the other entries in the table. There is a phantom date added to help but for tables where only two players are present it doesn't always help.
I am not sure what is going on with the QScores for Misotu's games. I spent a couple hours last weekend looking through the code and nothing obvious popped out.
The way EQM uses only one score per table per player doesn't seem to be represented by the QScore Query. I will dig into it some more this weekend. The problem is that I don't know it is broken. Scoring anomolies for sparsely populated tables is difficult to solve for.
Making significant changes to the scoring algorithm at this point is not wise. I don't want to turn the rankings upside down. If I can find a bug, I will fix it, but if it is working as designed and the problem is that with the sample sizes...
I am not sure what is going on with the QScores for Misotu's games. I spent a couple hours last weekend looking through the code and nothing obvious popped out.
The way EQM uses only one score per table per player doesn't seem to be represented by the QScore Query. I will dig into it some more this weekend. The problem is that I don't know it is broken. Scoring anomolies for sparsely populated tables is difficult to solve for.
Making significant changes to the scoring algorithm at this point is not wise. I don't want to turn the rankings upside down. If I can find a bug, I will fix it, but if it is working as designed and the problem is that with the sample sizes...
Similar threads
