

I have read that previous to the Deluxe edition, adding 3 influence to the US made victory distribution balanced. Sorry to raise again the topic of imbalance again, but i can´t find post referring specifically to Deluxe edition with optional cards
I know that the Deluxe edition changes (2 US influence in Canada, plus the optional cards) have removed part of the previous imbalance in victories between sides, butm, is the general opinion that the sides are currently balanced if playing deluxe with optional cards? (no china card variant)
We have recently hold a championsions using deluxe + optionals. CCCP get 38 victories versus 27 US victories. The numbers are still low enougth to be considered anecdotal, and also players with different experience were involved. Anyway, we were using a VP bid to choose sides, and though generally people bid for the CCCP 1 or 2 VP, there where people comfortable to take US for 0VP, both when offered to then, or directly choosing (we roll a die, best roll bids first for a side, opponet raises or acept offered side)
Anyone has more data about the ratio of victories? If so, is there anny consesous of which handicap should be assigned to a side to made the most balanced game?

Shawn Garbett
United States Nashville Tennessee
Will Provide Statistics for Data

zorro_jv wrote: CCCP get 38 victories versus 27 US victories. The numbers are still low enougth to be considered anecdotal
The 95% confidence interval range for the true population proportion for CCCP victory percentage with this championship data is 46.48% to 70.44%. This overlaps 50% so there is a lack of evidence to reject the null hypothesis that the true proportion is 50%. This confirms that at best the evidence is anecdotal.
Now with more data sets, one can combine them using metaanalysis.

Daniel Hogetoorn
Netherlands Amsterdam

On this page you can download the ETSL game results of season 1 and 2 to compare with your statistics: http://unitedfriendseu.wordpress.com/europeantwilightstrug...
In my opinion the DeLuxe game with optionals is completely balanced. Additional influence doesn't really help an unexperienced US player. It's mainly the different tactical challenges which make the USSR slightly stronger. An experienced USSR player will always defeat an unexperienced US player. But an unexperienced USSR player is able to defeat an experienced US player, since the Early War tactics for the Russians are quite easy to understand. That's why I think the Russians win more. If we have a look at the best of 3 series in both the ETSL and the Wargameroom league, most of them are decided in 2 games (so the best player wins, not the one playing Russia). And as you can see in the Excel sheet, almost all games have been played without or with max. 1 additional influence for US.

Riku Riekkinen
Finland Jyväskylä

Currently wargameroom ( I believe thoughest tourney that uses optionals) uses +1IP to US anhywhere it already has IPs ( overcontrol allowed ) ( = always Iran ). The stats still say USSR is favoured.

Daniel Hogetoorn
Netherlands Amsterdam

Riku Riekkinen wrote: Currently wargameroom ( I believe thoughest tourney that uses optionals) uses +1IP to US anhywhere it already has IPs ( overcontrol allowed ) ( = always Iran ). The stats still say USSR is favoured.
True, but in fact your records already prove that sides are not really the decisive factor. When a player wins 100% of his 20(!) matches so far (and playing US more often than USSR), skill is much more important. I checked the matches and saw a few top players playing USSR twice beating weaker US players. This gives the impression that the USSR is stronger, but I still think the balance issue is overrated.
But anyway, as I already said earlier, bidding is the fairest way to go. Who thinks playing USSR is worth 3 IP to the US will pay for it, especially against Riku, the best US player ever! ;)

Shawn Garbett
United States Nashville Tennessee
Will Provide Statistics for Data

DeDaan wrote:
Combining season 1 and 2, there is no evidence that season nor bid level influenced the outcome in a binomial model. Thus, we are left with a winning percentage proportion, which the USSR wins 48.88% of the matches with a 95% confidence interval of 42.31% to 55.43%. Thus there is no evidence in this data that the true value is not 50%.

Benoît Delcorps
Belgium Bruxelles

Personally, i think the deluxe version with optional cards is perfectly balanced...

Riku Riekkinen
Finland Jyväskylä

I claim that, if there would be only little need for agreeing need to raise the bid to +1 (from 0), it would have never happened. There are many people who believe every game is balanced out of the box or are just used to play that way.
Now also believing that it would drop back to 0 while the stats say the US is getting too little is also quite strange to me. The next vote will be +1 against +2.

Shawn Garbett
United States Nashville Tennessee
Will Provide Statistics for Data

Riku Riekkinen wrote: Now also believing that it would drop back to 0 while the stats say the US is getting too little is also quite strange to me.
The binomial model I just ran doesn't say this at all. It says that there is no evidence the bidding had an effect on outcome. With statistics, given more trials, it may say otherwise, but at present there's no indication that it has any impact at all on game outcome with this data set.

Riku Riekkinen
Finland Jyväskylä

CyberGarp wrote: Riku Riekkinen wrote: Now also believing that it would drop back to 0 while the stats say the US is getting too little is also quite strange to me. The binomial model I just ran doesn't say this at all. It says that there is no evidence the bidding had an effect on outcome. With statistics, given more trials, it may say otherwise, but at present there's no indication that it has any impact at all on game outcome with this data set.
Did you get your data from wargameroom or ETSL?

Shawn Garbett
United States Nashville Tennessee
Will Provide Statistics for Data

ETSL



CyberGarp wrote: Riku Riekkinen wrote: Now also believing that it would drop back to 0 while the stats say the US is getting too little is also quite strange to me. The binomial model I just ran doesn't say this at all. It says that there is no evidence the bidding had an effect on outcome. With statistics, given more trials, it may say otherwise, but at present there's no indication that it has any impact at all on game outcome with this data set.
Thanks for taking the time to do the calculations:thumbsup:. Although i understand what you say (or at least i think i do :D), im unable to calculate myself :)



Riku Riekkinen wrote: CyberGarp wrote: Riku Riekkinen wrote: Now also believing that it would drop back to 0 while the stats say the US is getting too little is also quite strange to me. The binomial model I just ran doesn't say this at all. It says that there is no evidence the bidding had an effect on outcome. With statistics, given more trials, it may say otherwise, but at present there's no indication that it has any impact at all on game outcome with this data set. Did you get your data from wargameroom or ETSL?
The info for wargamer room seems a bit confuding, at least to me... the table at the end of each league page (pages like this) is the compilation of all games played? if so, is the leftplayer the winner?
If so maybe we can get more data for CyberGarp (although previous editions are a bit more confusing, being used different balance options)

Riku Riekkinen
Finland Jyväskylä

Anyway.. what international big tournaments there are? Wargameroom league, World Boardgaming Championships, Ladder, ETSL... Now the current champions of all those leagues play in wargameroom. There was a vote for a balance and the current result is +1IP to US. So yes, there is that kind of agreement between good players. Now there was a lot of disagreement then also, so its only the current compromise.



zorro_jv wrote: the table at the end of each league page (pages like this) is the compilation of all games played? if so, is the leftplayer the winner?
If this supposition is true, this are the wins:
Winners,
Using only Optional Cards: URRS USA wargameroom 2012 80 73 wargameroom 2013 1 2 ETSL 1 34 34 ETSL 2 53 56 Using Optional Card + 1 IP for USA: wargameroom 2012 22 18 wargameroom 2013 160 130 ETSL 1 21 16 ETSL 2 43 42
subtotal, only optional 168 165 subtotal, optional +1 IP 246 206 TOTAL 414 371



The numbers are odd, or i get them totally wrong: the US wins less when they are given one additional IP!
That happen both on wargameroom and ETSL. I think that if Shaw can do the calculations, it will show that numbers are not enought to say that there is proven imbalance.



Riku Riekkinen wrote: Anyway.. what international big tournaments there are? Wargameroom league, World Boardgaming Championships, Ladder, ETSL... Now the current champions of all those leagues play in wargameroom. There was a vote for a balance and the current result is +1IP to US. So yes, there is that kind of agreement between good players. Now there was a lot of disagreement then also, so its only the current compromise.
It would be interesting to have acces to game result from a more general base apart from results for top/torunament players... It's seems that the game is generally well balanced (although is possible that there is a sligth imbalance that more experienced players can exploit)



If those numbers are right, it would appear the players can exploit the imbalance by giving the 'advantage' to your opponent. It may be noise in the data, but it may be a real effect: that extra 1 IP gives the US player the player a boost in confidence, they play a riskier game, and lose more often as a result. Curious.

Riku Riekkinen
Finland Jyväskylä

One thing to note from the stats is the qualitive side (opposed to quantitive). In 2012 top players mutually agreed during season that the game was not in balance. In wgr players can, if they both agree, change the bid before the play (before side are determined). So the +1 in 2012 which shows the USSR were still the advantaged side is probably played by the people who knew to use that option.

Shawn Garbett
United States Nashville Tennessee
Will Provide Statistics for Data

zorro_jv wrote: The numbers are odd, or i get them totally wrong: the US wins less when they are given one additional IP!
That happen both on wargameroom and ETSL. I think that if Shaw can do the calculations, it will show that numbers are not enought to say that there is proven imbalance.
The problem with statistics in this instance is one is doing proportional tests. A lot of data is required to answer any question with confidence. So, while the truth may be that the IP bid is quite influential, the number of games required to show that difference can be quite high. This is a critical point about statistics. There may be a real difference (and I personally believe there is), but there's not enough data to show it previously.
> prop.test(246, 452, p=0.5, alternative=c("greater"))
1sample proportions test with continuity correction
data: 246 out of 452, null probability 0.5 Xsquared = 3.365, df = 1, pvalue = 0.0333 alternative hypothesis: true p is greater than 0.5 95 percent confidence interval: 0.5044594 1.0000000 sample estimates: p 0.5442478
Here with the total data quoted above, one tests if the true proportion giving the US one IP, does the USSR have an advantage in this case. The number show that the answer is yes. From a statistical viewpoint, it's just barely significant. From a gamer perspective, a 5% edge is a 5% edge. Very odd. However, from a statistical viewpoint, the evidence is just on the edge. A counterintuitive result!!!
A twosided test is not significant.



Riku Riekkinen wrote: One thing to note from the stats is the qualitive side (opposed to quantitive). In 2012 top players mutually agreed during season that the game was not in balance. In wgr players can, if they both agree, change the bid before the play (before side are determined). So the +1 in 2012 which shows the USSR were still the advantaged side is probably played by the people who knew to use that option.
I don´t understand what you are pointing... I thought that maybe stronger veteran players recognize URSS superiority, and sio they bid to play with it. A noso veteran/strong players may accept the bid, playing with the US +1 IP. Then the URSS win, but because is palyed by the veteran/strong players.
But that does not explain why the disproportion is higher in wargameroom 2013, where i think there are no bids for side (you play once with each side, don´t you?).
Another posible explanation is that URSS AR1 coup in Iran is not the best possible opening for the URSS. (hey, i'm just theoric) If US get 1 more influence there, URSS is less prone to coup there, and more prone to open elsewhere, maybe with a better result. I know this sound odd... but not more odd than letting west germany empty playing the USA! That was previuosly seen as nosense, but now is more used.
Although i respect veteran players opinion (and share it: in my recent tournament i have to play with the URSS each time), maybe agreements are based on invalid data (as with the empty w. germany)... i would try to keep recovering data...



4100xpb wrote: If those numbers are right, it would appear the players can exploit the imbalance by giving the 'advantage' to your opponent. It may be noise in the data, but it may be a real effect: that extra 1 IP gives the US player the player a boost in confidence, they play a riskier game, and lose more often as a result. Curious.
Or it may force ( at least psychological ) USSR player to a more aggressive approach which proves to be successful ...

Daniel Hogetoorn
Netherlands Amsterdam

zorro_jv wrote: But that does not explain why the disproportion is higher in wargameroom 2013, where i think there are no bids for side (you play once with each side, don´t you?).
In the preliminary phase, all matches in the Wargameroom league are played with random sides. In the ETSL players almost always (unless a decisive match is needed) play an even amount of matches against eachother (both players play both sides once or twice versus each other).


