Some stats based on the Istari points from 1650 and 2950 scenarios.
Using the premise that victorious allegiances have higher Istari points per nation, we can presume a higher average Istari points rating for all nations in an allegiance, would indicate a greater win percentage for them.
I have noted the following from the posted Istari results for 1650 and 2950 nation. (As always, a word of caution on these stats, they only take into account a small amount of data, neutrals are not considered etc.)
In 1650 the average Istari rating for all nations per allegiance and their top 10 highest scores are:
FP - 1389 (1st place for Sinda + Noldo seemed incorrect and I used 1860 and 2050 respectively for them) DS - 1392 NE - 1587 (no data for Rhudaur, though this would most likely decrease this figure when available)
Between the FP and DS the figures look very even with a slight advantage to the neutrals which is to be expected.
In 2950 the average Istari rating for all nations per allegiance and their top 10 highest scores are:
FP - 995 DS - 1190 NE - 1280
As in 1650 the neutrals have a slight edge, however there is a notable difference between the FP and DS (~20%) and if this translated into victories then the DS would seem to start in this scenario with a substantial advantage.
It is difficult to quantify the advantage and the 20% mentioned above may be greater or less than the actual % games won by each allegiance, though I would be surprised if there was an even split between the DS and FP in 2950. I would be interested to hear any thoughts you may have on this or whether there is other considerations or data to be poured over.
The figures above were calculated from the top 10 nation results per allegiance as follows:
Allegiance nation scores were ordered in rank from 1 -> 10 place
Each place (1st -> 10th) was averaged for all nations (FP 1st places = 1697, 10th places = 1202)
The average of each placing was totalled and averaged in turn (1389)
Paul
PS if anyone wants to validate my excel sheets, just let me know as these were calculated quickly.
In the old GSI days the DS won two-thirds of the games. Your figures suggest a rough 50-50. However, one fly in the soup may be that LOSING DS nations appear in the Istari ratings. Those high scoring nations were defeated by FP One Ring victories.
I imagine Harley can produce win/loss numbers for games and nations.
Originally posted by 88 Noldo
[b]In the old GSI days the DS won two-thirds of the games. Your figures suggest a rough 50-50. However, one fly in the soup may be that LOSING DS nations appear in the Istari ratings. Those high scoring nations were defeated by FP One Ring victories.
I imagine Harley can produce win/loss numbers for games and nations. [/b]
Trying to use the top 10 scores to determine win % of alliances would be like using the average height of the 10 tallest trees in the world to try to figure out the age of the average tree on earth.
1650 games start ever few weeks. Say, 20 a year for 10 years. That would be 200 games. Even if there is a 66% DS win rate, that is 66 games the FP have won. The top 10 scores from the 660 winning positions are bound to be high.
All analys of DS to FP top 10s can show is how high people have scored with the various positions, before anyone was really trying to sore well.
Oh, at it can give goals to see if you can beat the score that people got before they were really trying.
It is a valid point regarding one allegiance losing more regularly with higher points distorting the figures.
Also in the olden days didn’t the DS win quite a bit more often before agents were downgraded a little?
I do not have ME Game stats and would love to see them, but then I wouldn’t have the pleasure of looking over the numbers.
Though perhaps I could work the tree analogy out from them.
Trying to use the top 10 scores to determine win % of alliances would be like using the average height of the 10 tallest trees in the world to try to figure out the age of the average tree on earth.
Darrell,
Using your logic, it is a coincidence then that the average VP scores for the top 3 nations per allegiance (identified below for 1650) would not be reflected by these nations winning more regularly than others with lower scores?
Using your logic, it is a coincidence then that the average VP scores for the top 3 nations per allegiance (identified below for 1650) would not be reflected by these nations winning more regularly than others with lower scores?
[/b]
Define “winning”. The Noldor finishing #1 in VPs when the FP win… Using that definition, then there is some validity to looking at Top 10 VPs for a nation.
However, you were trying to show that higher VPs for the DS mean they win more often. This is simply not the case.
Looking at the top 10, you’re looking at the flukes. It is like checking the age of the oldest trees on a contenant to determine the average age of trees on a continent.
I think your definition of “some validity” would equate to them definitely being the most regular top 3 nations using the VP system or how else would they have highest average VP scores?
But my point is NOT about whether VP’s are a valid measure of nation success, and because of your obsession with this you have missed my point, which is that winning allegiances would generally have better characters, armies, pop centers and more gold (higher VPs). I have not played in a One Ring victory game and I guess it could throw a spanner in the above assumption, though for the sake of simplicity we will presume the FP and DS can split One Ring wins.
In 2950 I have pointed out that the DS nations have a 20% higher average VP rating than the FP, your argument that they are flukes is based on what evidence, why are the top 10 nations results different from others? I expect we could draw the same conclusions from any sample of 10 games except the bottom 10 because of the artificially lowest score of 400.
You tree analogy does not stand up, obviously using 10 trees to determine anything about millions of other trees would be useless, but I am talking about 10 from the number of 2950 games completed in the last ~3 years (which I believe Clint has used). If you look at the front sheet the last 2950 game filled in 10 weeks which would equal to ~5 per year, (this should be higher since I am omitting grudge, variants and the fact more new players has led to games starting more regularly) so the top 10 results are much more likely to be a sample of 10 from 20-40, rather than the millions you anticipate.
The results and 20% difference may have no significance if we are likely dealing with a low number of games. A good run off victories for the DS could easily account for the disparity, One Ring victories could contribute, more experienced DS players or there could be an inherent advantage for the DS.
My point was to understand why there is a difference, if it was related to a DS advantage and if so is this advantage as significant as 20%, I continue to believe the DS are stronger at game start.
Let’s say the DS have won 150 games and the FP have won 75. That means that these “top 10s” are ignoring 140 DS victories and 65 FP victories.
All the “higher for the DS” shows is that the DS have clumped the points on a few nations more often than the FP have.
Again, you’re counting the oldest trees on each contenant, and trying to use it to judge the number of trees on each continent.
Well, if there are more trees, then they’d have more chance of getting older, right?
WRONG! Certain trees (positions) do better. Just because DS tend to group points slightly more than FP, doesn’t mean DS win more often.
Try this explaination. Many clai that time is on the DS side. That is, the longer the game, the more often the DS win. So, the FP victories tend to be quick games where the Woodmen never have a chance of scoring well against the Noldor, but in long games, the BS can score as well as the DkLts. Since we’re only looking at the top 10, we can’t begin to estimate the median, mean, etc.
Your analysis is flawed from conception up. It is utterly meaningless.
I do get your point and for a larger number of games I agree it would be flawed, however I disagree that we are dealing with 150 games and I expect we are dealing with +/-20, and that some of the lower top places of each allegiance would most likely be from losing sides.
Your explanation of the DS having more points would indicate that the FP stick it out longer in losing situations than the DS? Why would this happen except that they were naturally weak from the start and accepted this fact?
I would say that I am counting the 10 tallest trees of two tree species in a small garden and concluding one species is naturally taller than the other.
Perhaps we will see the 2950 win ration some day and see whether there is an even split or not.
Originally posted by klub I do get your point and for a larger number of games I agree it would be flawed, however I disagree that we are dealing with 150 games and I expect we are dealing with +/-20, and that some of the lower top places of each allegiance would most likely be from losing sides.
About 20 games? Harly starts a 1650 Open Game every couple weeks. 1650-17 would be on GT4 and 1650-35 is already going. That is 18 games in about 2.5 months, just 1650.
They said they used all the games they’ve run, plus info they got from DGE. What is that? 4 years of games?
The guy at the top of the “chess” style rating had to have won about 22 games more than he lost (or won a lot of games where the opponents had a lot more players).
No… I think we’re looking at more than 100 1650 games. Looking at the top 10 is meaningless.
Then again, before game 17 ended on GT4, all 4 of the positions I’ve completed were in the top 10 for the nation. Maybe it is only a couple months of games… No. One of the ones they listed was NKA2. That has to be a year ago. They also had from 1650-101 (ended 3/03)… But they were missing 2950-239 (ended 9/03).
When I complained that half my positions were missing, the other half went away.
I don’t know how many games are represented, but if the info is at all accurate (which it isn’t) then it SHOULD be more than 100-150 games.
Originally posted by 88 Noldo Darrell, are you saying Harley removed you from the PRS because you have complained about it?
How to answer that, while being politically correct…
I’ll just list the events.
Clint welcomed discussion on PRS.
I complained… A lot.
Clint started blocking my posts from the Yahoo list, not because they violated the rules, but because I was “spamming”.
PRS results were posted. (page 1s anyway)
I asked why I wasn’t listed with players that are 3 for 3…
“You have opted out.”
“Well opt me in.”
“I’m thinking of taking action against you for behavior detrimental to the game.”
“WHAAAAAT?”
“Okay, I’ve decided to not take action for or against you.”
I took this to mean he wasn’t opting me in.
Then, the additional pages started working. I pointed out that I was listed as 2 for 2, but I’m really 3 for 3. That only 2 of my 4 positions were listed in the positional top 10 (played 2 positions one game) but all 4 should be. That despite being 4 for 4 and having all 4 position finish in the top 10 by position… I was only 69th (or so) on the ranking that was supposed to measure your ability to score well in a position?
The result of me pointing out my missing data was that the half of my results that were orignially listed, went away.
Darrell I don’t appreciate you quoting from a private email. But to clear things up you originally asked to be removed from the PRS. I did so. Unfortunately one of the staff accidently put you on the list and then I removed you as per your earlier request.
Due to your spamming, abusive and aggressive attacks both out of game and also your comments that you would attack anyone with a higher rating than you on your own team. So yes you did violate the rules.
With no rating to check against - which is the reason you gave me for wanting your ratings - there is no-one to attack. So for that reasons I have not put you back on the ratings list.
As for you new PRS - I asked that you get 10 players to support you. I await that list.
I don’t think that you would particularly appreciate me posting or quoting emails from you to the public - please don’t do so from me.
