The role of luck in 7 wonders

[doctorchris]

I have been playing 7 wonders on BoardGameArena for about 9 months, and have started playing the new edition in the last month. At times my ELO has been well above 400, high enough to be ranked in the top 100 players. But it has always come down again quite rapidly. Currently my ELO is 317.

As far as I can tell, I don't play any differently over time. I always play with 7 players and choosing the wonder side. So I feel luck must play a big part in the variation in my ELO. Certainly the times I have been over 400, it only takes a couple of mediocre games to drop 50 points. The most obvious random elements seem to me to be which wonder you get (Hali good, Olympia bad), your access to resources nearby which can totally mess up construction of the wonder, how militaristic your neighbours turn out to be, and which guilds appear in your hand at the start of age 3.

Has anyone had any success at keeping their ELO score above 400 for an extended time? If so would you care to share how you overcome these random elements?

Or any other thoughts on the role of luck in 7 wonders more generally.

[Wonderful Plays]

Very interesting topic.

But before I comment on the luck element, I want to flip this argument on its head and ask: what is the behavior from the ladder and your score that you would expect? Infinite climb to the top? Win 1, loose 1, repeated?

What I'm trying to say here, this is a multiplayer game, swings are wider than in 1v1, just because you have more players at the table. At some point you reach a place, where it's hard to improve (like with any other game) and you'll have to really study to figure out the small advantages.

Back to the luck in 7 Wonders.

This is a game where you essentially make 18 moves (6 cards choices per 3 eras), system designed here is very tight and we have a lot of people on the ladder that played over couple thousand games. For a lot of games this would be way too much and the optimal moves might have been found a long time ago. Instead we established a metagame were there a lot of known strategies, but there is still room for creativity.

First I'd like you to tell me what player count do you play, because it does change a lot.
I'd say the most "stable" is 3 player, which is the easiest player count to play. You see each hand 2 times, it's easiest to visualize, the concept of just red and green being good is easy to understand. You can be stable there, but the room for innovation is the lowest. I'd say the wonder and distribution matters a lot. One thing to note, because of only 3 hands there is less variation in the card spread, which is good for stability, but again bad for creativity.
Then we have 4 player, my personal favorite player count, where you see 3rd and 4th hand only once and there is additional guy in the pool, that can in some games be a random element, but this is where you are actually playing the whole table the most.
Here I would add some more questions: When picking from a hand do you think about the person opposite of you? For example when playing against Halikarnassos do you prioritize early grays that they don't have access to? This particular line is aimed at making sure that you are in an advantegous strategic postion (not opposite to green) and usually forces them to take more red (which is almost always good for you).
For higher player counts it gets more chaotic, but appart from 7P, which in my opinion doesn't work that great, they can be very strategic as well. In 6p the game is flipped on its head, you can't wast energy fighting your neighbours, suddenly you try to make sure that all 3 of you have picks from hands so you'll try to pursue opposite strategies.

There are a couple of skills that can help you improve:
- hand tracking
- predicting your opponents moves from each hand
- visualizing left cards in the pool
- trying to put yourself in a strategic situation that will maximise your chances
- not always trying to win, realising when to play for 2nd or even 3rd sometimes

I'd point you to this guide that IMO started competitive 7 Wonders:
https://boardgamegeek.com/thread/691370 ... -7-wonders
And shamelessly gonna plug my channel:
https://www.youtube.com/watch?v=LWeNZYVPaew
In the video I show just how good Olympia can be in 4 player.

Finally: is there luck in 7 wonders? Yes. Some wonders are better than others, some distribution will be very easy to play, when others are almost impossible to win against other good players. But is there skill in maxisiming your chances and outcomes? Plenty. I've been able to reach 500 constantly, even when experimenting on the ladder (I do fall down because of it from time to time) and I've been able to even reach 600, when playing with much lower rated player at the time, I've climbed to 500 using 6P, 5P, 4P, 3P. After initial learning the game, the skills are very subtle, but they are there.

Cheers.

[doctorchris]

Thanks for your very thoughtful reply.

I always play 7 person. That's partly because the game is called 7 wonders, not 3 wonders or 4 wonders. But mainly becuase when I play IRL it will usually be with 7 players, as there are so many other good games for 3 and 4 players, but very few for 7 where you don't have to wait for ages for others to make their plays. So I would like to develop skills that are going to be most useful IRL. Your suggestions are very helpful.

I guess that means I will have to put up with more randomness than in a 3 or 4 player game. Maybe I should try going down to 5 or 6 player and see how I get on.

[Wonderful Plays]

7 players is the worst player count by far.

1. You can't see one hand, for some wonders this will mean never being able to build certain part of the wonder.
2. There is a shortage of brown resources 1,14 single per player, 0,86 dual per player, 5,14 brown resource per player (only worse in 3p, but when someone build in 3p everyone has access to it).
3. There are 3 carravanseries and people with them will have much better time in the game due to point 2.
4. Every green is doubled, that mean that you are at a mercy of card distribution much more than any other player count. This promotes already strong Hali B even more.
5. Every wonder is in the game and that's bad. A lot of the times this wonders will have the same plan and there is less shifts in strategy. The game is called 7 wonders, because there were 7 wonders in the ancient world, not because you have to play it 7 player.

I really suggest trying 3P, that's the easiest to learn the most competitive variants, and dive into 4P or even 5P/6P. 7P is just there, because it kinda works, but honestly it propably could have not been on the box even.

[doctorchris]

Very good points about Caravanseries, and not even seeing one hand. Sometimes that last hand has had both the double resources I needed and I don't get a chance to even see them. I'll try some 5 or 6 player games and see how I get on.

[dschingis27]

At times my ELO has been well above 400, high enough to be ranked in the top 100 players. But it has always come down again quite rapidly.

Adding to this: It is very common for 7Wonders that your Elo goes up and down in an interval that spans about 100 points. It happens also to the very best players, e.g. the current Arena champion's Elo as well as the Elo of the champion from Season 2 go up and down roughly between 400 and 500.

[ucarts]

In my opinion, much of the issue is in the rating system rather than the game itself (although I'd love a version of the game with better balance). I've had the same problem of 50 or 100-point rating swings in a short period of time; it's inherent to the rating system. This is a real problem in a game where 100 rating points can make the difference between being a top player and not even being ranked.

I would really love the system to be tweaked to reduce the Elo change for a single game. Right now, one's rating can go up or down 20-30 points from one game. That might be appropriate for players who have only played a handful of games, but for experienced players, it should really be closer to 5 points.

I don't have much to add in terms of strategy. From my experience, it's much harder to maintain rating when playing against lower-rated players in 7p than in 3p. As more players are added to the game, there's less you can do to predict which cards are coming your way, and your score is influenced more by what your neighbors do. For example, if both of your neighbors are building heavy military, it's difficult to score higher than the players with neighbors that are not building military. It wasn't too hard for me to get to 400 in 3p; I would have a much worse time doing so in 7p.

[Storm Giant]

As preceding reply, said, it's more of rating system than the game itself, (or more specifically, the number of players). I only play 7P game and have come to realize that ratings (from different player counts) are not meaningfully comparable. A 400 rating gained from playing only 3p is not comparable to a 400 rating gained from only playing 7p. The less number of opponents, the easier it is to win, and vice versa. When you win in a 3p game, you've defeated 2 other players; in 7p, you've defeated 6 other players (3 times more). For that reason, 7p is the most challenging - the most difficult - player count to win in - consistently. Of course, the definition of 'consistently' is up for debate. If you are one of the top players for 3p, 'winning consistently' might be winning about 9 games out of 10, and for top 4p, it might be 8-9 games out of 10. For 7p, it's way lower, maybe about 3 out of 10 games (about 30 wins per 100 completed games).

OP mentioned reaching 400+ ratings (accross many many games) by playing only 7P. That's really impressive. There aren't many players who can do that. I myself, got past 400 but fell below and having not an easy time getting back to it.

Right now there is no really easy way to determine how good a player is on different player counts (well, only for 3p and 4p, with arena, and ELO rating). But for higher player counts, there's nothing (ELO not suitable). Reaching 400 rating periodically, is a good indication as well as having a higher win% but there's no ready data available for comparison.

As for tip on 7p, aside from those given above: tracking key cards. If you are going science, then track the green cards that you have not seen yet (or might be coming your way). If you are not going science, track the military cards, especially those you haven't seen yet (easier with the new ed now, as there are 2 copies of each card in 7p). Last tip: Just play because you enjoy playing it.

[doctorchris]

Thanks to you all for your thoughts. I took the advice and have been playing 5P for a couple of weeks. My ELO went down to 225 while learning it, but has recovered a bit. Now I'm going to bite the bullet and learn 4P, which several of you have said is the best. I know my ELO will take another hit, but as Storm Giant said, I'll be playing because I'm enjoying learning something new.

[tap-zbyna]

7 players is the worst player count by far.

You made some very good points and basically I agree with you.

But I do not egree with you on 3p and 4p games. 3 or 4 players makes 7w very much of a luck driven game. Caravansery is an essential card. There is only 1 caravansery card in those games. I would like to see a statistics of winning 7w game without a caravansery not using the mass green strategy. I think the number would be very very low for 3 or 4 players game. In 3 or 4 players wonders like Ephesos or Alexandria can easily make a monopoly with their base resource.

Also in 4 players the green strategy is much too easy to counter. You can bury green cards very easily (particullary school).

I like 5 or 6 players game the most, the luck is not as important as in 3,4, and 7 wonders.

In 3 or 4 players Halikarnassos can be completely useless - especially with Babylon-B at the table - since there is not much to pick from with your wonder ability. On the other hand Rhodos or Alexandria are bit of overpowered in 3p game. Rhodos can control battles easily and Alexandria has caravansery by default. If Alexandria gets the second Caravansery, it is almost unstoppable in 3rd age.