Elo Rankings

[rbodleyscott]

The PBEM server code that was supposed to calculate Elo ranking updates was not working correctly, hence all ranked games completed up till now will not have been scored. (The ranking page showed "maintenance", because there was no data).

It has now been fixed for Ancients, so ranked games completed from now on should get ranked correctly, and the adjusted rankings shown on the ranking page.

https://www.slitherine.com/leaderboard/ ... f-glory-ii

Sorry it has taken us so long to figure this out.

Medieval has not yet been fixed, but we hope to fix it ASAP.

[Witan]

I just won 4 battles against vanilorus in Ancient FoG2, but the ranking page shows him with 1060 and me with 940. Does that mean that now a lower number is a better sign?

[rbodleyscott]

I just won 4 battles against vanilorus in Ancient FoG2, but the ranking page shows him with 1060 and me with 940. Does that mean that now a lower number is a better sign?

How did you manage to win 4 games in the 80 minutes since the previous problem was fixed? Or did you actually win them previously when the system was bugged?

Obviously the winner should get the higher score, so that would be a bug. But when I tested it by conceding a game with rbodleyscott vs robdleyscott2, rbodleyscott2 was correctly awarded the higher score.

[Witan]

I just won 4 battles against vanilorus in Ancient FoG2, but the ranking page shows him with 1060 and me with 940. Does that mean that now a lower number is a better sign?

How did you manage to win 4 games in the hour since the previous problem was fixed? Or did you actually win them previously when the system was bugged?

Obviously the winner should get the higher score, so that would be a bug. But when I tested it by conceding a game with rbodleyscott vs robdleyscott2, rbodleyscott2 was correctly awarded the higher score.

Of course these battles started some days ago but all 4 battles were won by me just minutes after you did this post here, but the ranking page shows a wrong score at least in my case.

[rbodleyscott]

I just won 4 battles against vanilorus in Ancient FoG2, but the ranking page shows him with 1060 and me with 940. Does that mean that now a lower number is a better sign?

How did you manage to win 4 games in the hour since the previous problem was fixed? Or did you actually win them previously when the system was bugged?

Obviously the winner should get the higher score, so that would be a bug. But when I tested it by conceding a game with rbodleyscott vs robdleyscott2, rbodleyscott2 was correctly awarded the higher score.

Of course these battles started some days ago but all 4 battles were won by me just minutes after you did this post here.

Oh dear. Back to the drawing board then!

[rbodleyscott]

To help us track down the problem, did you win all 4 games by achieving the normal victory conditions, or did vanilorus concede, was the turn limit reached, or what?

(I lost my test game by conceding, and that was scored correctly)

[Witan]

To help us track down the problem, did you win all 4 games by achieving the normal victory conditions, or did vanilorus concede, was the turn limit reached, or what?

I won all 4 quite decisive by the regular victory conditions.

[Witan]

Jorgito78 now also confirmed to me that he in fact lost the battle against heidipie, but still got a score of 1015 while heidipie only has 985 despite winning.

[rbodleyscott]

Jorgito78 now also confirmed to me that he in fact lost the battle against heidipie, but still got a score of 1015 while heidipie only has 985 despite winning.

Thanks.

Can you list the rout %ages in any of the games?

[rbodleyscott]

It is fairly clear that the PBEM code for determining Elo rank changes is not working as intended. We are assisting the PBEM programmer to fix the issue(s).

(It was tested in the open beta test, but if it wasn't adjusting the scores correctly, nobody pointed it out)

Once the code is working as intended, we will probably need to reset all of the scores to 1000. (Because correcting all the errors might be prone to error).

[Witan]

Jorgito78 now also confirmed to me that he in fact lost the battle against heidipie, but still got a score of 1015 while heidipie only has 985 despite winning.

Thanks.

Can you list the rout %ages in any of the games?

Yes of course. In the first one that ended I won with 7% losses as Carthaginians while the Galatians lost 48%. In the mirror battle I won with 18% losses as Galatians while vanilorus lost with 43% as Carthaginians.

In another battle I won with 4% losses as Georgians while the Seleucids lost 48%. In the mirror I won with 13% losses as Seleucids while the Georgians lost 49%.

[Witan]

It is fairly clear that the PBEM code for determining Elo rank changes is not working as intended. We are assisting the PBEM programmer to fix the issue(s).

(It was tested in the open beta test, but if it wasn't adjusting the scores correctly, nobody pointed it out)

I also took part in the Open Beta of Medieval, but the Score worked fine for me there. Sadly i cannot tell regarding FoG2 Ancients, because I had no time to participate there.

[SnuggleBunnies]

As far as I recall the elo score adjustment *did* work correctly during the Medieval beta.

[tyronec]

As far as I recall the elo score adjustment *did* work correctly during the Medieval beta.

Likewise for all the games I playtested.

[rbodleyscott]

While we are working on this, please report all incorrect ranking adjustments AND ALSO correct ranking adjustments, with both games. (And say which game, and give both user names)

(Note that the two games have different PBEM code bases, so they may or may not both be bugged).

[tyronec]

Hi Richard,
Do you know if this has been fixed or if Slitherine are working on it ?
Just to know so that players will be motivated to use the system in player created games.

[Karvon]

From staff comments on Discord, they are working on it, but no prediction of how soon it'll be fixed. They will be wiping the ELO rankings and start fresh once the issues are resolved.

[rbodleyscott]

From staff comments on Discord, they are working on it, but no prediction of how soon it'll be fixed. They will be wiping the ELO rankings and start fresh once the issues are resolved.

Correct. It is proving problematic to fix it without risking breaking something else. It might not get done till the New year.

[raultow]

I just play and win a game of a mirror macht. When I came to see the ELO chart, for my surpise, my oponet had more ELO rating!

[rbodleyscott]

I just play and win a game of a mirror macht. When I came to see the ELO chart, for my surpise, my oponet had more ELO rating!

This is a known issue. It is being worked on.