Streaks and patterns?

[Hattusilis_III]

Does anyone else find that they get ridiculous streaks of bad luck?*
Every time - without fail (I have been logging the details of my 2000 games) - when I start my first gaming session of the day I simply struggle to win a game and typically lose about 7 or 8 out of ten. Most of these seem to be to low ELO opponents so my ELO often drops by about 100 or so in that first session. Then I gradually work my way back up to a high ELO again and the same thing happens the next day? Within just a few days my ELO has moved between 400 and 100. Often those losses are to opponents scoring 15 or 20 above their average scores or I just get a series of times that don't score much (eg: one series of 5 games produced 13 navigators and 1 cartographer for zero points; a separate game saw 5 cartographers no navigators for 5 pts while my opponent had 3 cartographers and 3 navigators for 20 points).

This might sound like a whinge but it's really not meant to be. I really enjoy the game and have played against a lot of really good and fun opponents. I have learnt a lot by watching better players' moves. I also recognise there is a fair bit of randomness in the game and have no problem with random extreme outcomes like I mentioned above. Crazy outcomes are also really fun IRL as you can laugh about it with your friends (but not so much online).**

However, the weird streaks are frustrating and seem more than just random, but rather patterns or streaks. It's like when you choose a random board but usually you play the same board for 3-4 turns in a row before switching. That's not truly random. Or the number of times when the start of a game either virtually mirrors a previous game or your opponent's tiles. I don't play nearly as much IRL but I never see those repeat patterns.

I'm just curious if anyone else experiences this. I gather some of the really high ranking players don't because I rarely see their ELO drop really low but I do see others with wide-ranging ELOs.



* Yes - I recognise that nearly everyone only ever complains about their bad luck without recognising their good luck. However I have logged my games and included all data for certain streaks and have found that the extent and patterns of bad luck are much greater and more persistent than the extent of good luck in winning streaks.


** Here are some examples I can think of:

My opponent draws 4 gunners in the first 4 tiles, places them all, and then proceeds to draw 3 straight monkeys.

On a separate occasion I get 3 gunners in first 4 tiles and no monkeys rest of the game. Only about 10 games later I get 3 gunners in first 3 tiles and no monkeys rest of the game again. In both cases the very first tile I draw in the game immediately following is... you guessed it, a monkey.

My opponent scores 18 points on their final tile to beat me on a tie (they are holding the map).

These are hysterical, chaotic fun IRL, but on BGA they just make you go hmmm...

[Jellby]

Does anyone else find that they get ridiculous streaks of bad luck?*

Tons of users. The forums are littered with "game is rigged" and "dice are not random" type of posts. You can join the ranks.

It's like when you choose a random board but usually you play the same board for 3-4 turns in a row before switching. That's not truly random.

You're wrong. That is (or may well be) truly random. ABABAB is as random as AAABBB or AABBBA or AAAAAA. You want alternating, not random.

These are hysterical, chaotic fun IRL, but on BGA they just make you go hmmm...

Probably because you play a lot more on BGA than IRL, you have more personal interactions IRL (that take away the focus from what actually happens in the game), and you don't care about elo when you play IRL, I presume.

There might be bugs, wrong coding logic, or hacks/exploits, I think we have seen cases of all at some point (in other games), but it takes a lot more than a feeling and a bunch of cherry-picked "bad luck" cases to make anyone look further into it.

[Hattusilis_III]

I'm not suggesting that anyone look into it or that it is rigged, I was just asking if anyone else experienced the same thing. I understand probability and perception.
Logging and reviewing the data from 2000+ games isn't "cherry picking".
Frequent and recurring sequences of game boards isn't necessarily random, like frequent and recurring series of coin tosses or dice rolls are not.
I suspected someone would engage combatively rather than simply discussing.

[Hattusilis_III]

So to simplify, the question is just, do other people experience these dramatic swings in fortune with this game?

[Meeplelowda]

This game is swingy and there's only so much you can do about it. The current peak ELO is 505. Compare it with the peak ELO in a game with 0 randomness. Those games have peak ELOs in the 800s or 900s, and one I found over 1000. When there's no randomness, the skill edge will always dominate. When randomness is involved, the skill edge gives you an advantage, but won't eliminate the variance.

[Hattusilis_III]

This game is swingy and there's only so much you can do about it. The current peak ELO is 505. Compare it with the peak ELO in a game with 0 randomness. Those games have peak ELOs in the 800s or 900s, and one I found over 1000. When there's no randomness, the skill edge will always dominate. When randomness is involved, the skill edge gives you an advantage, but won't eliminate the variance.

Yeah, I definitely think all of that is correct.

My observations suggest that most people experience big swings. My suspicion is that some aspects of the random-generating component may exacerbate that.

[Meeplelowda]

My suspicion is that some aspects of the random-generating component may exacerbate that.

Then I'll refer you to the dozens of other "something's fishy about the RNG because streaks like this never happen with real dice/cards" threads. In particular, look at the one in the backgammon sub-forum where someone claimed something NEVER happens with real dice and I provided the log from an IRL backgammon tournament where it in fact did happen.

[Hattusilis_III]

My suspicion is that some aspects of the random-generating component may exacerbate that.

Then I'll refer you to the dozens of other "something's fishy about the RNG because streaks like this never happen with real dice/cards" threads. In particular, look at the one in the backgammon sub-forum where someone claimed something NEVER happens with real dice and I provided the log from an IRL backgammon tournament where it in fact did happen.

Yes I am aware of the problem with "stranger than fiction" perception and the concepts of probability.
However, I wouldn't have posted if I didn't feel that the repeat patterns were nevertheless still out of the ordinary as I'm not talking about a strange event in a single game but patterns and streaks across multiple games.

[Meeplelowda]

My suspicion is that some aspects of the random-generating component may exacerbate that.

Then I'll refer you to the dozens of other "something's fishy about the RNG because streaks like this never happen with real dice/cards" threads. In particular, look at the one in the backgammon sub-forum where someone claimed something NEVER happens with real dice and I provided the log from an IRL backgammon tournament where it in fact did happen.

Yes I am aware of the problem with "stranger than fiction" perception and the concepts of probability.
However, I wouldn't have posted if I didn't feel that the repeat patterns were nevertheless still out of the ordinary as I'm not talking about a strange event in a single game but patterns and streaks across multiple games.

Have you run your data through any statistical tool other than the "eye test" (to borrow a sports term)?

[Hattusilis_III]

Then I'll refer you to the dozens of other "something's fishy about the RNG because streaks like this never happen with real dice/cards" threads. In particular, look at the one in the backgammon sub-forum where someone claimed something NEVER happens with real dice and I provided the log from an IRL backgammon tournament where it in fact did happen.

Yes I am aware of the problem with "stranger than fiction" perception and the concepts of probability.
However, I wouldn't have posted if I didn't feel that the repeat patterns were nevertheless still out of the ordinary as I'm not talking about a strange event in a single game but patterns and streaks across multiple games.

Have you run your data through any statistical tool other than the "eye test" (to borrow a sports term)?

I've done neither. I have just done various calculations in Excel.