Hi guys.

I was also collecting some stats from the site, refreshed them today. They are not available online, but I'm sharing some data below. What is important about the data:

• all data comes from finished games. If there was drop, game was cancelled or simply finished before round 6, its not considered at all
• I have no information if game was with auction system or not, so bid values are also not considered. It basically means that average/max scores are lowered by games with auction system as I can't undo the bidding
• there is no additional games filtering (like not taking under account games with bad players, low results, etc)

So lets start with the most basic data, ordered by number of players and then by average scores

Faction	     Players      Max Score   Avg Score   Games
ivits	        4            220       149,97      582
itars	        4            226       149,69      613
ambas	        4            255       148,4       532
taklons	        4            222       146,62      541
firaks	        4            202       145,09      360
terrans	        4            239       144,28      727
baltaks	        4            206       143,61      453
gleens	        4            231       140,13      351
nevlas	        4            222       138,97      418
xenos	        4            204       138,41      478
geodens	        4            224       137,12      441
lantids	        4            202       135,05      299
hadsch-hallas	4            216       134,79      495
bescods	        4            205       133,72      358

ivits           3            254       168,73      304
ambas           3            237       157,4       267
firaks          3            232       155,78      186
itars           3            250       154,66      284
taklons         3            246       152,56      229
nevlas          3            231       151,84      210
terrans         3            223       151,43      371
baltaks         3            204       148,91      196
bescods         3            205       147,79      178
hadsch-hallas   3            220       147,28      258
geodens         3            226       145,59      244
gleens          3            220       144,32      186
xenos           3            208       139,69      200
lantids         3            195       134,64      124

ivits           2            251       176,17      716
itars           2            237       159,99      501
firaks          2            228       158,54      336
ambas           2            235       155,06      626
baltaks         2            215       151,6       349
nevlas          2            254       151,41      352
terrans         2            255       150,73      724
geodens         2            233       150,35      458
taklons         2            240       149,43      263
hadsch-hallas   2            224       147,21      442
xenos           2            227       145,23      316
bescods         2            211       143,86      252
gleens          2            210       143,15      326
lantids         2            213       129,77      115

So in all categories (number of players) Ivits has the higher average score, while in 2 and 3 player their domination is very noticeable. In 4 player game they are very close to Itars and Ambas. On other side we have Lantids performing really badly in 2-3 players games. What surprised me the most are Firaks average scores.

Average scores not necessary goes toe to toe with wins ratios as some of the games were with auctions system. And it also depends on other players results:

Players	Faction		Games	1st	2nd	3rd	4th	1st %	2nd %	3rd %	4th %
4	taklons		541	191	121	105	124	35.30%	22.37%	19.41%	22.92%
4	ivits		582	191	150	146	95	32.82%	25.77%	25.09%	16.32%
4	itars		613	194	171	125	123	31.65%	27.90%	20.39%	20.07%
4	ambas		532	168	136	119	109	31.58%	25.56%	22.37%	20.49%
4	terrans		727	202	179	183	163	27.79%	24.62%	25.17%	22.42%
4	firaks		360	93	88	92	87	25.83%	24.44%	25.56%	24.17%
4	nevlas		418	93	108	100	117	22.25%	25.84%	23.92%	27.99%
4	gleens		351	77	99	89	86	21.94%	28.21%	25.36%	24.50%
4	baltaks		453	96	135	111	111	21.19%	29.80%	24.50%	24.50%
4	xenos		478	101	118	143	116	21.13%	24.69%	29.92%	24.27%
4	lantids		299	60	66	81	92	20.07%	22.07%	27.09%	30.77%
4	geodens		441	88	105	126	122	19.95%	23.81%	28.57%	27.66%
4	bescods		358	66	74	101	117	18.44%	20.67%	28.21%	32.68%
4	hadsch-hallas	495	81	104	139	171	16.36%	21.01%	28.08%	34.55%

3	ivits		304	173	76	55	0	56.91%	25.00%	18.09%
3	taklons		229	95	67	67	0	41.48%	29.26%	29.26%
3	ambas		267	97	93	77	0	36.33%	34.83%	28.84%
3	nevlas		210	76	63	71	0	36.19%	30.00%	33.81%
3	firaks		186	67	65	54	0	36.02%	34.95%	29.03%
3	itars		284	102	91	91	0	35.92%	32.04%	32.04%
3	bescods		178	62	44	72	0	34.83%	24.72%	40.45%
3	terrans		371	120	134	117	0	32.35%	36.12%	31.54%
3	geodens		244	73	73	98	0	29.92%	29.92%	40.16%
3	hadsch-hallas	258	77	98	83	0	29.84%	37.98%	32.17%
3	baltaks		196	53	73	70	0	27.04%	37.24%	35.71%
3	gleens		186	49	67	70	0	26.34%	36.02%	37.63%
3	xenos		200	42	76	82	0	21.00%	38.00%	41.00%
3	lantids		124	23	40	61	0	18.55%	32.26%	49.19%

2	ivits		716	545	171	0	0	76.12%	23.88%
2	itars		501	267	234	0	0	53.29%	46.71%
2	taklons		263	140	123	0	0	53.23%	46.77%
2	firaks		336	170	166	0	0	50.60%	49.40%
2	ambas		626	314	312	0	0	50.16%	49.84%
2	nevlas		352	170	182	0	0	48.30%	51.70%
2	baltaks		349	166	183	0	0	47.56%	52.44%
2	geodens		458	217	241	0	0	47.38%	52.62%
2	terrans		724	330	394	0	0	45.58%	54.42%
2	hadsch-hallas	442	197	245	0	0	44.57%	55.43%
2	xenos		316	130	186	0	0	41.14%	58.86%
2	gleens		326	130	196	0	0	39.88%	60.12%
2	bescods		252	95	157	0	0	37.70%	62.30%
2	lantids		115	36	79	0	0	31.30%	68.70%

Domination of Ivits in 2-3 players games are also visible in their wins-ratio. In 4 players games Taklons are the best. However if you consider 1st and 2nd place together, then Taklons are third (57.67 %) while Ivits second (58.59 %) after Itars first (59.54 %). So its easier to win with Taklons, but by average better result you will get with Ivits and Itars. Lantids are the worst in 2-3 players games, and they do below average in 4 players games. Bescods are last but one in all player setups.

What conclusions we may get from above data:

• Game is more balanced in 4 players setups as the differences between scores and wins ratios are smaller
• Ivits are overpowered, especially in 2-3 players games
• Lantinds and Bescods are underpowered, especially in 2-3 players games. They are also the least picked factions
• Terrans are the most popular faction in all modes, however faction performs "only" above average

And cherry on the cake at the end - top 20 players with the best average scores (games with drops are not counted) with at least 10 finished games:

Players	Name			Games	Avg
4	turkishvancat		11	185,55
4	Rudeiro			11	182,27
4	IRIS			38	177,61
4	ZeroCool		23	177,48
4	Nolegsmoo		17	173,24
4	kyte			26	171,15
4	satchisuta		24	170,71
4	sorpy			17	170,24
4	Molfo			66	169,26
4	ssbon			25	169,16
4	lucasrrr		31	168,87
4	Spooky			26	168,85
4	DDi_yong		12	168,33
4	Wy			10	167,3
4	Babbuc49		99	167,18
4	Ewan			34	167,12
4	jgunter			14	167
4	Kesterer		25	166,88
4	Tennessee Walker	19	166,84
4	Stylish			10	166,8

3	EoMoR			23	188,91
3	Spin			14	187,93
3	Milkleo			16	181,5
3	Lemondas n.lin		18	180,11
3	mbouzada		10	179,9
3	Dinogau			15	177,07
3	gunsung			10	177
3	sorpy			10	176,6
3	seki			10	176,6
3	Akiyama1103		13	176,38
3	Babbuc49		22	174,23
3	Roadrunner19		12	172,42
3	Saltara			10	171,9
3	doldole			14	171,43
3	tanios			12	170,58
3	hikary27		30	170,43
3	TearsInTheRain		79	170,39
3	NineTailCat		10	170
3	redarox			27	169,96
3	Adriano			15	167,2

2	Aillas			14	196,64
2	testrun			49	191,53
2	bigpaullai		14	189,86
2	Amek			11	188,45
2	Milkleo			10	186,8
2	hikary27		18	185
2	MasN			12	184,58
2	b_			12	184,5
2	Akiyama1103		136	183,9
2	theyellow		71	183,45
2	Poziom			14	181,71
2	Ewan			11	181,45
2	Ansun			17	180,82
2	Gondalf			19	178,79
2	Kesterer		13	178,77
2	Adriano2		47	177,85
2	Babbuc49		30	177,07
2	naname			10	176,7
2	Molfo			12	176,58
2	seki			14	176,57

@coyotte508 - Is it possible to add bidding value to each player result in /api/game/closed endpoint? That would help to recalculate some stats related to average, max score values :)

Enjoy the data.

@AntoineBR In your example there is chance you or any other player will end up with Itars with 10-20 points bid on them, when some player decide to pick last faction - Terrans with 0 bid on them and immediately starts the game. And the player who end up with Itars doesn't have any opportunity to change their bid anymore. In your mode whenever some player decides to pick last faction, game starts right away.

@Molfo That way of playing was existing in Terra Mystica for quite long time, so there was no reason to not implemented it in Project Gaia. It just a mode, you are not forced to use it. If you don't like auction games, don't join rooms with auction mode enabled or create your own room. Simple as that. I don't see any reason to reduce available options how to play the game.

So far I've played maybe 10 games with auction mode, most of my PG games were in standard player order picks. What I can say is that it is definitely designed towards experienced players. Because, as you mentioned, its almost impossible to correctly estamite exact points difference between factions in particular game. Especially for players which are not experienced, they will probably bid towards factions the are familiar with or they simply don't see small oportunities hidden somewhere on the map or tech positions. So definitely bidding is part of the player skill and probably that's why experienced players prefer this mode over standard one. As @lucasrrr have said - auctions also makes less popular factions playable. You have opportunity to pick them and actually compete for 1st/2nd place. In standard mode it's rarely the case - for instance Lantids requires very specific setup to work well.

What I agree with you is that auction system shouldn't be used as a data input towards faction balancing. Unless you will just take scores from those games without bids values and calculate average score for each faction from all of the games. Anyway, faction balance is separate topic.

Barrage! With expansion as well. The only game close to Gaia complexity and replayability atm.

@AntoineBR It is cool idea but also has downsides. If you don't know which other factions will be picked, you can't correctly estimate the power of your faction in that particular game. You need to know in advance about your neighbours, their terraform distance, potential starting positions, gaia formers capacity, etc. Otheriwse you can't bid correctly. Image bidding for Itars which were picked first. It may appear as a great faction for the game, but if someone would picked Terrans as their last faction, Itars power dramatically drops as they gets another gaia planets competitor and very close neighbour. So last players has a big advantage as he has full game information and can correctly estimate which faction is the best for the game.

Anyway, no matter in which system factions will be picked, it never will be perfect. Bidding seems to be the most fair as each player has an opportunity to pick the best faction. And of course, it is annoying when someone picks bad faction for bidding. And maybe it is not fun to play such a faction as you know you won't get a good score, but still it is possible to win when you start the game with 20-30 points advantage.

I don't think so there is a better way to implement auction system. At least from the competitive point of view. In this system you have full game information, you know all the factions and their turn order. So entire bidding process is fair without favorizing any of the players. The main concern people have about this mode is that sometimes there is a faction in the pool no one wants to play with, which basically prolonging entire bidding process and forces one player to eventually pick it. But it is not competitive concern, its a fun concern. If that would happen during tournament, my guess is no would complain. Maybe even someone would be very happy about that pick and could think to himself sth like 'I think I can do quite well with that faction with this setup, lets see how much bidding I can force on other faction first. Lets hope I'm the only one thinking this way'. And after if he picked it and won with, it would be a fully deserve win.

So the question really is - what do players prefer? To have more competitive game or just have more fun with the game? It seems that we already have both. I don't mind if sth new will came up, like @AntoineBR idea for auction system. It is not as fair as current system, but it basically will force player to pick only good factions. So definitely we can have a new mode, but we also should keep the current one and let players pick what they prefer. However if tournaments will get introduced - we should use current auction system there.