Stats on factions

On https://forum.boardgamers.space/topic/82/statistics-just-wondering-on-the-progress/8 I've uploaded a script (which I hope is correct) that can fetch statistics for the factions on all games. For easy reference, here is a screenshot of the results:

Some thought, some of which I don't think will surprise experienced players:

• the Ivits are very strong overall, especially at 2 and 3 players
• the Taklons are very strong at 4 players, at which they have the most opportunities to rotate their brain stone
• the Nevlas also profit from higher player counts, and thus more rotation
• the Lantids and Gleens are weak at every player count
• the Itars, which are generally considered to be very strong, don't over-perform by that much
• the Bescods are very weak at 3 players
• average scores tend to go down as the player count goes up
• there doesn't seem to be a 'vanilla' faction that performs average on all player counts

I'm curious to hear what other players think!

Good job, thanks! Btw, in my opinion, these results show not mathematical but psychological regularities:

Most people like to win with as little effort as possible and don't like to lose. That is why they choose the most banal and easy factions, such as Terrans and Ambas Therefore, I cannot agree with your theses, especially with:

'the Lantids and Gleens are weak at every player count'

they are not weak factions - but the most of players can't use them well, because they usually choose Terrans (instead Lantids) and Xenos (instead Gleens) - why? Because most players want to win in the most easy way.

I also think that in order to have more meaningful stastistics we’ll have to wait that elo settles a bit and then filter out games involving one or more inexperienced players

Also auction games should be filtered out for this purpose i guess

Thanks a lot for the effort, this is great. I agree though that we will need more games and more options to filter via elo once that has settled.

Still interesting data already! But we´ll have to read it properly: in this game much more then in Terra Mystica (especially played exclusively on the standard map) basically evey faction probably has a setup which is good for it. Just some have way more of those scenarios. Still probably also Gleens and Lantids are viable.

This is very interesting, I was hoping for some stats to come out.

I run statistics on the games me and my friends have played in the last two years, and results are similar, as far as best and worse factions are concerned. I agree that Itars in particular are a very good faction but they do not overperfom (unlike the Taklons that, contrary to the opinion of many, really are one of the top factions in GP).

I find the "in between" factions more puzzling, I'm really surprised at seeing Bescods performing better thatn Hadschallas, Xenos and Baltaks, which I've always considered to be good factions (maybe not top-tier, but able to put up some good games). Also, Nevlas have a very high winning rate here, while they have average performances at best in my gaming group.

(All of the above is referred to 4p games, which I think are the most representative).

I agree with Babbuc, however, that a filter is needed. Games where players have dropped should not be factored. Also, it would proabably be better to take out every game where someone scored below 110. Based on my experience over hundreds of games, I'd say that's the lowest possible score for somebody who has experience with Gaia Project (again, in 4p games at least). Lower scores could be the result of non-experienced players joining the game, and this could bias the statistics.

Ariallito's point is also worth considering (to an extent: I think there's no denying that lantids and gleens really are in need of some help).

[EDIT] sorry, I just figured out that you've probably already taken out games where people have dropped

Great work! very interesting job! It can be implemented with win-rate from the statistical relationship between race/type of final1-2 objectives

Nice! Not surprised to see the Taklons perform well with four players.

Am surprised by the Bescod weakness with 3 players though (you would think, with 3 players, they'd be able to get their hands on more of their own planets than with any other player count, as you'd play with the full amount of sectors, but with less competition than with 4 players.)

@Molfo said in Stats on factions:

Ariallito's point is also worth considering (to an extent: I think there's no denying that lantids and gleens really are in need of some help).

You folks just need to figure out how to properly dolphin :P

@El-Temblo show the world how it’s done!

@Babbuc49

Ahah, will do :D

Seriously, GP has been out for 3 years now; data wojld be interesting, but there's general consensus that Itars and Ambas are OP, while Gleens and Lantids would benefit from a boost.

Hope the next expansion will fix this.

I'm not sure I would say that Itars and Ambas are OP (not without including Taklons in the lot, at least).

I agree on the fact that gleens and lantids are UP in most contexts, however. I don't think Itars and Ambas need to be nerfed, but a few other factions could benefit from a small buff to get on level (this includes the Bescods, that imho suffer from the same limitations as gleens an the dolphins).

If you want to win the game by playing Ambas, Taklons or Ivits, Terrans you have to use your brain 25-50%, you can drink a lot while playing, and it still shoukld be easy. But if you want to win by using Gleens or Dolphins, you have to put the vodka down and start thinking. This is Gaia Project's charm, so the expansion should NOT fix it.

I'm not sure i agree that Itars, Ambas and Ivits are absolute no-brainers. Maybe it's just me who don't play these 3 specific factions a lot, but I have to put in a lot of thinking when using them (while my Taklon, Terran or Firak games are much more streamlined). But then again, this might be a matter of being more or less used to play a certain factions.

The real strenght of Ambas, Taklons and Itars is that they are very versatile. They adapt well to a very high number of scenarios. This is why they get picked a lot (and often perform well, I'm not denying all these races have powerful abilities).

Other factions (e.g. Geodens, Firaks) are also good, imho, but can be picked in a more restricted number of settings. In the right context, however, they can be as powerful as the others.

I'm just not sure the same can be said about gleens, bescods and lantids. The last two in particular I've seen perform abysmally even in settings where they should have theoretically done well. But as I said, I'm not sure about this, would gladly hear other opinions.

Except Ambas, Taklons, etc. perform even better when the person piloting them is actually thinking. In terms of how much thinking each faction requires: Lantids - I'll give it to you Lantids require some thought on PI, ability, placement, etc. Gleens - On the other hand, for the most part its brainless. Your expansion is so constrained you rely more on the opponents to not be paying attention than actually playing well. On the other hand, the factions you mentioned that require 25%-50% brain capacity on the other hand: Terrans - Have multiple viable openers and requires power management, like any gaiaforming faction. Ambas - I admit are fairly fixed opening-wise in their strategy. Ivits - Have a ton of exclusive openings that you have to pay attention to, and you also have to pay attention to where you take planets/space stations. Also, they gaiaform about half the time, which means you have to make that decision and manage power well if you gaiaform. Taklons - Require the most power management of any non-gaiaforming faction, and if the other players don't mindlessly charge you, you have to make those chase decisions.

Honestly speaking, the more versatile factions like Taklons require more thought than more linear factions like Gleens, since you actually have to choose what you're doing.

@MasN said in Stats on factions:

Gleens - On the other hand, for the most part its brainless. Your expansion is so constrained you rely more on the opponents to not be paying attention than actually playing well.

lol. yes, i mostly agree with this

in general, yes, what you said, i probably have to focus less when playing taklons because i played them a lot, while ambas for instance not so much.

i 100% agree in particular that ivits require a lot of planning ahead.

in general, though, thes kind of things are also heavily dependent on the player and their specific styles of play and preferences

Oh wow, didn't want to start an argument about the balance of factions 😅; mine were the thoughts I've collected across the bgg board - but please, take into account that James Ataei himself admitted that Ambas have been intentionally powered up because of a bias which occurred during playtesting (https://boardgamegeek.com/thread/2213655/why-do-itars-generate-3-knowledge-their-academy/page/2). And someone has yet to explain me convincingly why Itars start with an extra ore, and why their AC1 generates one more knowledge.

@El-Temblo i also came to the conclusion that ambas seem to overperform when using the map balancing app, but i’ve been thinking lately that in a vacuum they’re not as strong as itars or taklons, and also i’ve been thinking that for those two they really aren’t that much OP either once players at the table are mindful of what they’re letting them have (leech, tracks, techs) and most of the time they CAN be denied stuff effectively. That said i’m looking forward to see what the general opinion on faction balance will be in a year or so, now that we’re going to be able to properly collect huge amount of data with online games

@Babbuc49 erhm... the balancing algorithm issue is addressed by James in the very thread I linked :P

I agree on your conclusion about the two factions you mention, though: Taklons can easily perform a mad economy, and Itars have the two (inexplicable) advantages I mentioned ON TOP OF the already super strong faction+PI abilities they already have. Actually, Itars are so strong that you can (and often will) win without bothering to open their OP AC1 :)

Still, it looks like there is just no reason for which Ambas produce 6 extra ores across the course of the game, and that makes me mad :D.

In general, I concur with the opinions already expressed here about the mentioned factions.

Let me be clear, though: given the complexity of the game mechanism, I LOVE Gaia Project, and think that the authors, developers and playtesters have carried out a miraculous work of balancing it all. It just looks like there could be a few small tweaks which would make the game even... more perfect :).

A few final words about a "weak" faction: Lantids. Ah, my beloved Lantids... they look VERY difficult to balance. They have an incredibly strong and unique ability, which absolutely needed to be tuned down a bit in order not to make them uncontrollable. Still, I wonder why -2 starting c and no t income were chosen as their "nerfs"... I'm dolphinly curious about why these decisions were made. Ideas?

about itars: i think their 3k academy is ok because it's really hard to use effectively that AND their PI ability in the same game... so it's actually a very versatile faction because you can either open ac and get lots of "k"s from early on, or you can open PI and start burning power to gain techs early on. Both options are viable, and give rise to totally different games, which adds variety to the game.

As for their extra starting ore: yeah that makes no sense at all and should just be removed. It's the only nerf to any faction on which I'd agree, I think.

About the Lantids, I think the idea of giving them very little power was to compensate their ability to form federations very easily. This however turned into a major nerf because it also means not being able to access power actions as easily as other factions, especially in the early game. The two lesst starting credits are just overkill imho, and makes it nearly impossible to open AC with them, which is what normally dissuades me to pick them.

NB: to be clear, yes, you can build an academy in R1 with lantids, with the right booster, but at a very high price, which imo is not worth paying.

I suspect the -2c for the Lantids might be about making the PI, 4+ mines opening (my favorite) a little harder to execute (3+6+8 = 17c would be easily obtainable with the right booster).

@ariallito only began playing this game about 6 months ago live, and a couple months now on this app, but I think Gleens can be pretty good and relatively easy to play, just under relatively narrow circumstances. You need at least two of three: gia round bonus 3, 4, or end game gia bonus. Also necessary that the 3pt gia card isn’t stuck in a bad spot, and there are a good number of desert and gia planets within two jumps of each other. When those conditions are met, I’d say maybe 1/4 of the time(?), you can pretty easily get 175+ pts in 3-4 player games. 5pts per gia with 8-9 pts per a couple rounds is pretty strong!