Terrible algorithms

[Gulchen]

@ BackGa ​ :



No (due to the "not" in your question). ​ It's just, more games make it more likely that unusual things happen.


I imagine that was mostly bad luck: ​ Backgammon has a large amount of luck, so the
ranks are probably much closer together in elo than they would be for other games.

(I don't, but maybe: ​ Do you have an idea of how many loses
against equal-elo opponents it would take to go from 50th to 950th?)

[Jellby]

Yesterday I went from top 50 to 950th, pure bad luck?

And maybe some sub-optimal play from your side. What else? Are you suggesting some pixie inside BGA's servers is secretly targetting you to make you lose? Even if it were a bug (of which there is absolutely no indication at this point), it would be bad luck that it affects you more than your opponents.

[veggivet]

Many bg players on this site find the algo to be exceedingly streaky. This is just one of the methods used to 'addict' you to keep playing, IMHO, and it is not unique to this site.

[Remkar]

Many bg players on this site find the algo to be exceedingly streaky. This is just one of the methods used to 'addict' you to keep playing, IMHO, and it is not unique to this site.

You sound like someone who may not understand how BGA, the random function, and individual game development on this site work.

The people who make these games are not BGA employees (I suppose there could be a small number of exceptions, though). They are usually volunteers completely unaffiliated with BGA, or occasionally hired by game publishers to implement their game on here.

The random function is a general function ALL BGA games can call on, and it doesn't know or respond differently based on what game is calling it, and it does not know the current game state of any individual game that is calling the random function.

It would be MUCH more difficult to add code to a game to try to cheat players than to just make the game.

The only way what you say could be true is if there is a big conspiracy between BGA employees and volunteer programmers involving a lot of people keeping secrets and doing LOTS of extra coding for no real benefit.

Additionally, it is known that people are not good judges of randomness (and many people with advanced education and decades of experience in statistics, psychology, and computer programming have posted as much on here).

I also realize that people who want to believe in conspiracy theories aren't likely to be swayed by anything, but maybe somebody out there reading this will learn something they didn't know.

Also, if anything I said above is inaccurate, I'll gladly edit my post, so please let me know.

[Jellby]

Many bg players on this site find the algo to be exceedingly streaky. This is just one of the methods used to 'addict' you to keep playing, IMHO, and it is not unique to this site.

Yes, that's exactly how randomness behaves, and why people are "addicted" to any king of random thing, from casinos to lottos to horse races to stock exchange.

If it's a conspiracy, it's much higher than BGA

[smattathias]

There's a lot of focus on sore losers complaining about bad rolls. So I'd like to flip that script and see how it lands.

It SUCKS to win a game with an unhealthy amount of luck. It sucks all the fun out of the game. Playing better, making better moves... None of it matters if it comes down to highly unusual dice behavior.

I've wondered how much of that is due to a lack of stakes, ie people play until the end because the stakes don't matter. Because of that, the luck of a good or bad roll plays a greater role.

Doubles account for 6 of 36 possible rolls. So it's strange to see several doubles in a row, several games in a row, several days and weeks and months in a row. Very odd.

[Remkar]

Doubles account for 6 of 36 possible rolls. So it's strange to see several doubles in a row, several games in a row, several days and weeks and months in a row. Very odd.

But see, that's what we're saying, it's NOT odd!

6 out of 36 (or 1 in 6 after reducing), is not super low.
And with the number of dice rolls in a typical backgammon game, we would EXPECT back to back doubles 1 or 2 times every match as a baseline.


And regarding your other comment about luck... Backgammon is a game with LOTS of luck. Once people reach a reasonable skill level, luck becomes a MUCH bigger factor than skill in most games. If you're looking for a game where luck doesn't impact you so much, there are plenty of other games with little to no luck involved.

[1skeezicks]

Doubles seem crazy. Just finished a game where opponent rolled double 6s four times in a row...seems fairly odd. BTW, I also think I roll doubles more frequently than normal during the course of a game. It's not unusual for me to get (9) doubles.

[Remkar]

Doubles seem crazy. Just finished a game where opponent rolled double 6s four times in a row...seems fairly odd. BTW, I also think I roll doubles more frequently than normal during the course of a game. It's not unusual for me to get (9) doubles.

Your lifetime stats show that you actually roll slightly less doubles than expected (it takes you almost 6.1 rolls per double instead of the expected 6).

Regarding an individual run of luck, you can't judge that out of the context of a huge number of rolls. There are 10s of thousands (100s of thousands, millions?) of backgammon rules every day on this sight. We would expect crazy strings if doubles to happen to each day. The 99.99% of games it doesn't happen in don't come on here to post about it, though.

[euklid314]

Doubles seem crazy. Just finished a game where opponent rolled double 6s four times in a row...seems fairly odd. BTW, I also think I roll doubles more frequently than normal during the course of a game. It's not unusual for me to get (9) doubles.

You have already played 346 games of Backgammon with (on average) 25 moves for you and 25 moves for your opponents. The probability that your opponent, or you, or both of you together, rolled four double-6s in succession is approximately 2%.

In other words: Out of 50 players that have played 350 games like you, one has experienced the 4 successive double-6s. As we have hundreds of players here we would have several of such posts a day if everybody were to post only this special occurrence in the forum.

Others could post if they see the successive sequence 1-1, 2-2, 3-3, 4-4, 5-5, 6-6 in their game. There are hundreds of "crazy" sequences that the human brain instantly recognizes and remembers. I am quite sure, however, that you cannot tell me how often you had the sequence 1-3, 2-6, 4-4, 2-5, 1-2 in exactly this order. Probably never or perhaps you had it already 1 or 2 times. Some user on BGA will almost certainly roll it tomorrow but I doubt that we will read about it in the forum - although its probability is less than your case of four double-6s.